wip align naming

wip fix suggestion
✨(summary) improve speaker assignment
2026-07-27 04:09:26 +00:00 · 2026-05-13 17:14:38 +02:00 · 2026-05-13 17:12:29 +02:00 · 2026-05-12 16:58:07 +02:00 · 2026-05-12 16:26:16 +02:00 · 2026-05-12 16:26:16 +02:00
86 changed files with 6419 additions and 5859 deletions
@@ -150,13 +150,14 @@ jobs:
        uses: actions/setup-python@v6
        with:
          python-version: "3.13"
-          cache: "pip"
-      - name: Install development dependencies
-        run: pip install --user .[dev]
+      - name: Install uv
+        uses: astral-sh/setup-uv@v7
+      - name: Install the project
+        run: uv sync --locked --all-extras
      - name: Check code formatting with ruff
-        run: ~/.local/bin/ruff format . --diff
+        run: uv run ruff format . --diff
      - name: Lint code with ruff
-        run: ~/.local/bin/ruff check .
+        run: uv run ruff check .

  lint-summary:
    runs-on: ubuntu-latest
@@ -322,6 +323,11 @@ jobs:
      - name: Checkout repository
        uses: actions/checkout@v6

+      - name: Install ffmpeg
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y ffmpeg
+
      - name: Install Python
        uses: actions/setup-python@v6
        with:
@@ -8,6 +8,32 @@ and this project adheres to

 ## [Unreleased]

+### Added
+
+- 🔒️(backend) add validation of Room.configuration
+- ✨(helm) add support multiple transcribe worker / endpoint #1247
+- ✨(backend) make LiveKit Egress recording encoding configurable #1288
+- ✨(summary) add speaker-to-participant assignment
+
+### Changed
+
+- ♻️(summary) change tasks endpoint signature
+- ⬆️(dependencies) update urllib3 to v2.7.0 [SECURITY]
+
+### Changed
+
+- 🧑‍💻(agents) use `uv` for package management
+- ✨(summary) improve speaker-to-participant assignment
+
+### Fixed
+
+- ♻(frontend) standardize role terminology across localizations
+- 🐛(backend) make start-recording atomic and fault-tolerant
+- 🔒️(frontend) room ids are generated with non-cryptographic rand
+- ⬆️(mail) fix dependencies not having resolved or integrity field #1321
+- 🐛(summary) complete webm support #1328
+- 🐛(backend) add link to "Open" text in recording email
+
 ## [1.15.0] - 2026-04-30

 ### Added
@@ -45,6 +71,7 @@ and this project adheres to
 - ✨(summary) allow more file extensions #1265
 - ♿️(frontend) refocus reactions toolbar with ctrl+shift+e is activated #1262
 - ♿️(frontend) set an explicit document title on recording download page #1261
+- ♿️(frontend) add customizable accessibility fonts #1270

 ### Fixed

@@ -109,7 +109,7 @@ build-frontend: ## build the frontend container
 .PHONY: build-frontend

 build-agents: ## build the multi-user-transcriber agent container
-	@$(COMPOSE) build multi-user-transcriber
+	@$(COMPOSE) build multi-user-transcriber-dev
 .PHONY: build-agents

 down: ## stop and remove containers, networks, images, and volumes
@@ -138,7 +138,7 @@ run-agents: ## start the multi-user-transcriber agent
 .PHONY: run-agents

 run-agent-multi-user-transcriber: ## start the LiveKit agents (multi users transcriber)
-	@$(COMPOSE) up --force-recreate -d multi-user-transcriber
+	@$(COMPOSE) up --force-recreate -d multi-user-transcriber-dev
 .PHONY: run-agent-multi-user-transcriber

 run-agent-metadata-collector: ## start the LiveKit agents (metadata collector)
@@ -109,7 +109,7 @@ k8s_resource('meet-backend', resource_deps=['postgresql', 'minio', 'redis', 'liv
 k8s_resource('meet-celery-backend', resource_deps=['redis'])
 k8s_resource('meet-celery-summarize', resource_deps=['redis'])
 k8s_resource('meet-celery-summary-backend', resource_deps=['redis'])
-k8s_resource('meet-celery-transcribe', resource_deps=['redis'])
+k8s_resource('meet-celery-transcribe-default', resource_deps=['redis'])
 k8s_resource('meet-backend-migrate', resource_deps=['meet-backend'])
 k8s_resource('livekit-livekit-server', resource_deps=['redis'])
 k8s_resource('livekit-livekit-server-test-connection', resource_deps=['livekit-livekit-server'])
@@ -101,6 +101,12 @@ update_npm_version "mail"
 # Update backend pyproject.toml
 update_python_version "backend"

+# Run uv lock in backend
+print_info "Running uv lock in backend..."
+cd "src/backend"
+uv lock
+cd -
+
 # Update summary pyproject.toml
 update_python_version "summary"

@@ -149,6 +155,7 @@ echo "      - src/frontend/package.json"
 echo "      - src/sdk/package.json"
 echo "      - src/mail/package.json"
 echo "      - src/backend/pyproject.toml"
+echo "      - src/backend/uv.lock"
 echo "      - src/summary/pyproject.toml"
 echo "      - src/agents/pyproject.toml"
 echo "      - CHANGELOG.md"
@@ -249,6 +249,7 @@ services:
  metadata-collector-dev:
    build:
      context: ./src/agents
+      target: development
    command: ["python", "metadata_collector.py", "dev"]
    environment:
      - LIVEKIT_URL=ws://livekit:7880
@@ -261,6 +262,7 @@ services:
      - AWS_S3_SECURE_ACCESS=False
    volumes:
      - ./src/agents:/app
+      - /app/.venv
    depends_on:
      - livekit
      - minio
@@ -269,6 +271,16 @@ services:
        - action: rebuild
          path: ./src/agents

+  multi-user-transcriber-dev:
+    build:
+      context: ./src/agents
+      target: development
+    env_file:
+      - env.d/development/multi_user_transcriber
+    volumes:
+      - ./src/agents:/app
+      - /app/.venv
+
  redis-summary:
    image: redis
    ports:
@@ -330,14 +342,6 @@ services:
        - action: rebuild
          path: ./src/summary

-  multi-user-transcriber:
-    build:
-      context: ./src/agents
-    env_file:
-      - env.d/development/multi_user_transcriber
-    volumes:
-      - ./src/agents:/app
-
 networks:
  default:
  resource-server:
@@ -100,6 +100,13 @@ sequenceDiagram
 | **RECORDING_STORAGE_EVENT_TOKEN**       | Secret/File | `None`                                                                                                                                                             | Token used to authenticate storage webhook requests, if `RECORDING_ENABLE_STORAGE_EVENT_AUTH` is enabled.                                                                                                                                                                                          |
 | **RECORDING_EXPIRATION_DAYS**           | Integer     | `None`                                                                                                                                                             | Number of days before recordings expire. Should match bucket lifecycle policy. Set to `None` for no expiration.                                                                                                                                                                                    |
 | **RECORDING_MAX_DURATION**              | Integer     | `None`                                                                                                                                                             | Maximum duration of a recording in milliseconds. Must be synced with the LiveKit Egress configuration. Set to None for unlimited duration. When the maximum duration is reached, the recording is automatically stopped and saved, and the user is prompted in the frontend with an alert message. |
+| **RECORDING_ENCODING_ENABLED**          | Boolean     | `False`                                                                                                                                                            | When `False`, LiveKit Egress uses its built-in `H264_720P_30` preset. When `True`, the `RECORDING_ENCODING_*` values below are sent to LiveKit as advanced `EncodingOptions`. See [Tuning recording encoding](#tuning-recording-encoding).                                                          |
+| **RECORDING_ENCODING_WIDTH**            | Integer     | `1280`                                                                                                                                                             | Recording video width in pixels. Only applied when `RECORDING_ENCODING_ENABLED` is `True`.                                                                                                                                                                                                         |
+| **RECORDING_ENCODING_HEIGHT**           | Integer     | `720`                                                                                                                                                              | Recording video height in pixels. Only applied when `RECORDING_ENCODING_ENABLED` is `True`.                                                                                                                                                                                                        |
+| **RECORDING_ENCODING_FRAMERATE**        | Integer     | `30`                                                                                                                                                               | Recording video framerate (fps). Directly impacts egress worker CPU (roughly linear). Only applied when `RECORDING_ENCODING_ENABLED` is `True`.                                                                                                                                                    |
+| **RECORDING_ENCODING_VIDEO_BITRATE_KBPS** | Integer   | `3000`                                                                                                                                                             | H.264 MAIN video bitrate in kbps. Only applied when `RECORDING_ENCODING_ENABLED` is `True`.                                                                                                                                                                                                        |
+| **RECORDING_ENCODING_AUDIO_BITRATE_KBPS** | Integer   | `128`                                                                                                                                                              | AAC audio bitrate in kbps. Only applied when `RECORDING_ENCODING_ENABLED` is `True`.                                                                                                                                                                                                               |
+| **RECORDING_ENCODING_KEY_FRAME_INTERVAL_S** | Float   | `4.0`                                                                                                                                                              | Keyframe interval in seconds. Drives seek granularity in the recorded MP4 (a player can only seek to keyframe boundaries). Larger values give the encoder slightly more bits for non-keyframe content at a fixed bitrate. `4.0` is a standard VOD value. Only applied when `RECORDING_ENCODING_ENABLED` is `True`. |


 ### Manual Storage Webhook
@@ -141,3 +148,54 @@ Using default project meet

 This allows you to verify which recordings are in progress, troubleshoot egress issues, and confirm that recordings are being processed correctly.

+## Tuning recording encoding
+
+By default, LiveKit Egress records with the built-in `H264_720P_30` preset: 1280×720 at 30 fps, 3000 kbps H.264 MAIN video and 128 kbps AAC audio. For a one-hour meeting this produces a file of roughly **1.4 GB**, which is often heavier than necessary for talking-head content and screen sharing.
+
+The `RECORDING_ENCODING_*` settings let operators override this preset without modifying the source. Values are passed straight through LiveKit's `EncodingOptions.advanced` to the GStreamer pipeline (`x264enc` for video, `faac` for audio), so there are no hidden conversions — what you set is what the encoder receives.
+
+### How values map to GStreamer
+
+| Setting                               | GStreamer element | Property                           |
+| ------------------------------------- | ----------------- | ---------------------------------- |
+| `RECORDING_ENCODING_WIDTH/HEIGHT`     | capsfilter        | `video/x-raw,width=W,height=H`     |
+| `RECORDING_ENCODING_FRAMERATE`        | capsfilter        | `framerate=F/1`                    |
+| `RECORDING_ENCODING_VIDEO_BITRATE_KBPS` | `x264enc`       | `bitrate=kbps` (kilobits)          |
+| `RECORDING_ENCODING_KEY_FRAME_INTERVAL_S` | `x264enc`     | `key-int-max = interval × fps`     |
+| `RECORDING_ENCODING_AUDIO_BITRATE_KBPS` | `faac`          | `bitrate = kbps × 1000` (bits)     |
+
+The H.264 profile is fixed to MAIN and the x264 `speed-preset` to `veryfast` by LiveKit (real-time constraint) — lowering the framerate is therefore the main lever to save CPU, while lowering the bitrate is the main lever to shrink the output file.
+
+### Reference profiles
+
+Rough 30-minute file-size estimates assume video + audio bitrate multiplied by duration. Actual sizes vary with content (static talking heads compress better than heavy screen motion). Egress CPU figures are indicative, measured on a single Ryzen laptop core saturated by the default preset (= 100 %); scaling is roughly linear with `framerate × bitrate` but the absolute numbers depend on the host hardware.
+
+| Profile                | Resolution | FPS | Video (kbps) | Audio (kbps) | Keyframe (s) | ~ size / 30 min | Egress CPU (vs. default) | Suitable for                                        |
+| ---------------------- | ---------- | --- | ------------ | ------------ | ------------ | --------------- | ------------------------ | --------------------------------------------------- |
+| Default (preset)       | 1280×720   | 30  | 3000         | 128          | 4            | **~690 MB**     | 100 %                    | Unchanged LiveKit behaviour                         |
+| Balanced               | 1280×720   | 20  | 1000         | 96           | 4            | ~240 MB         | ~67 %                    | Mixed content, moderate motion                      |
+| **Low CPU / small file** | 1280×720 | 15  | 600          | 64           | 4            | **~150 MB**     | ~50 %                    | Talking-head dominant meetings + occasional slides ★ |
+| Slide-heavy            | 1280×720   | 15  | 900          | 64           | 4            | ~210 MB         | ~55 %                    | Frequent dense screen sharing (decks, IDE, docs)    |
+| Minimum CPU            | 960×540    | 15  | 500          | 64           | 4            | ~125 MB         | ~30 %                    | Voice-first meetings, readable text not required    |
+| Audio-heavy fallback   | 1280×720   | 10  | 400          | 96           | 4            | ~110 MB         | ~35 %                    | Long webinars, low motion                           |
+
+★ Recommended starting point for typical LaSuite Meet usage.
+
+Environment variables for the **Low CPU / small file** profile:
+
+```bash
+RECORDING_ENCODING_ENABLED=True
+RECORDING_ENCODING_WIDTH=1280
+RECORDING_ENCODING_HEIGHT=720
+RECORDING_ENCODING_FRAMERATE=15
+RECORDING_ENCODING_VIDEO_BITRATE_KBPS=600
+RECORDING_ENCODING_AUDIO_BITRATE_KBPS=64
+RECORDING_ENCODING_KEY_FRAME_INTERVAL_S=4.0
+```
+
+### Caveats
+
+- **Screen-share readability — think bits/frame, not bitrate**: at 720p, text legibility starts to break down below ~40 kbits/frame (= `bitrate ÷ framerate`). The recommended preset (600 kbps × 15 fps) sits at exactly that threshold, comfortable for talking heads with occasional slide sharing. The same 600 kbps at 30 fps would only deliver 20 kbits/frame and visibly blur dense slides — which is why **lowering framerate is a more screen-share-friendly lever than lowering bitrate**. For deck-heavy or IDE-share meetings, prefer the **Slide-heavy** profile (900 kbps × 15 fps ≈ 60 kbits/frame).
+- **Motion handling**: the `veryfast` x264 preset is set by LiveKit and cannot be overridden here. Low-bitrate settings will therefore show more artefacts on fast motion than an offline re-encode with a slower preset would. This is the other reason FPS reduction is the safer tuning lever for meeting recordings.
+- **Audio**: AAC at 64 kbps stereo is transparent for voice but starts to compress music noticeably. Keep 128 kbps if you expect music playback in meetings.
+- **Codec choice**: H.264 MAIN is hardcoded on purpose. Switching to HEVC or VP9 would increase egress CPU cost 2×–5×, defeating the goal of this tuning.
@@ -68,6 +68,18 @@ SUMMARY_SERVICE_ENDPOINT=http://app-summary-dev:8000/api/v1/tasks/
 SUMMARY_SERVICE_API_TOKEN=password
 RECORDING_DOWNLOAD_BASE_URL=http://localhost:3000/recording

+# Recording encoding (LiveKit Egress advanced options).
+# When RECORDING_ENCODING_ENABLED is False (default), LiveKit uses its built-in
+# H264_720P_30 preset (1280x720, 30fps, 3000 kbps). Enable and tune to reduce
+# file size and CPU load on the egress worker.
+# RECORDING_ENCODING_ENABLED=False
+# RECORDING_ENCODING_WIDTH=1280
+# RECORDING_ENCODING_HEIGHT=720
+# RECORDING_ENCODING_FRAMERATE=30
+# RECORDING_ENCODING_VIDEO_BITRATE_KBPS=3000
+# RECORDING_ENCODING_AUDIO_BITRATE_KBPS=128
+# RECORDING_ENCODING_KEY_FRAME_INTERVAL_S=4.0
+
 # Telephony
 ROOM_TELEPHONY_ENABLED=True

@@ -19,7 +19,7 @@
        "@types/office-runtime": "^1.0.35",
        "acorn": "^8.11.3",
        "babel-loader": "^9.1.3",
-        "copy-webpack-plugin": "^12.0.2",
+        "copy-webpack-plugin": "^14.0.0",
        "eslint-plugin-office-addins": "^4.0.3",
        "file-loader": "^6.2.0",
        "html-loader": "^5.0.0",
@@ -36,7 +36,7 @@
        "source-map-loader": "^5.0.0",
        "webpack": "^5.95.0",
        "webpack-cli": "^5.1.4",
-        "webpack-dev-server": "5.1.0"
+        "webpack-dev-server": "5.2.1"
      }
    },
    "node_modules/@apidevtools/json-schema-ref-parser": {
@@ -4319,44 +4319,6 @@
        "node": ">= 4.0.0"
      }
    },
-    "node_modules/@nodelib/fs.scandir": {
-      "version": "2.1.5",
-      "resolved": "https://registry.npmjs.org/@nodelib/fs.scandir/-/fs.scandir-2.1.5.tgz",
-      "integrity": "sha512-vq24Bq3ym5HEQm2NKCr3yXDwjc7vTsEThRDnkp2DK9p1uqLR+DHurm/NOTo0KG7HYHU7eppKZj3MyqYuMBf62g==",
-      "dev": true,
-      "license": "MIT",
-      "dependencies": {
-        "@nodelib/fs.stat": "2.0.5",
-        "run-parallel": "^1.1.9"
-      },
-      "engines": {
-        "node": ">= 8"
-      }
-    },
-    "node_modules/@nodelib/fs.stat": {
-      "version": "2.0.5",
-      "resolved": "https://registry.npmjs.org/@nodelib/fs.stat/-/fs.stat-2.0.5.tgz",
-      "integrity": "sha512-RkhPPp2zrqDAQA/2jNhnztcPAlv64XdhIp7a7454A5ovI7Bukxgt7MX7udwAu3zg1DcpPU0rz3VV1SeaqvY4+A==",
-      "dev": true,
-      "license": "MIT",
-      "engines": {
-        "node": ">= 8"
-      }
-    },
-    "node_modules/@nodelib/fs.walk": {
-      "version": "1.2.8",
-      "resolved": "https://registry.npmjs.org/@nodelib/fs.walk/-/fs.walk-1.2.8.tgz",
-      "integrity": "sha512-oGB+UxlgWcgQkgwo8GcEGwemoTFt3FIO9ababBmaGwXIoBKZ+GTy0pP185beGg7Llih/NSHSV2XAs1lnznocSg==",
-      "dev": true,
-      "license": "MIT",
-      "dependencies": {
-        "@nodelib/fs.scandir": "2.1.5",
-        "fastq": "^1.6.0"
-      },
-      "engines": {
-        "node": ">= 8"
-      }
-    },
    "node_modules/@pkgr/core": {
      "version": "0.2.9",
      "resolved": "https://registry.npmjs.org/@pkgr/core/-/core-0.2.9.tgz",
@@ -4370,19 +4332,6 @@
        "url": "https://opencollective.com/pkgr"
      }
    },
-    "node_modules/@sindresorhus/merge-streams": {
-      "version": "2.3.0",
-      "resolved": "https://registry.npmjs.org/@sindresorhus/merge-streams/-/merge-streams-2.3.0.tgz",
-      "integrity": "sha512-LtoMMhxAlorcGhmFYI+LhPgbPZCkgP6ra1YL604EeF6U98pLlQ3iWIGMdWSC+vWmPBWBNgmDBAhnAobLROJmwg==",
-      "dev": true,
-      "license": "MIT",
-      "engines": {
-        "node": ">=18"
-      },
-      "funding": {
-        "url": "https://github.com/sponsors/sindresorhus"
-      }
-    },
    "node_modules/@types/body-parser": {
      "version": "1.19.6",
      "resolved": "https://registry.npmjs.org/@types/body-parser/-/body-parser-1.19.6.tgz",
@@ -6686,21 +6635,20 @@
      "license": "MIT"
    },
    "node_modules/copy-webpack-plugin": {
-      "version": "12.0.2",
-      "resolved": "https://registry.npmjs.org/copy-webpack-plugin/-/copy-webpack-plugin-12.0.2.tgz",
-      "integrity": "sha512-SNwdBeHyII+rWvee/bTnAYyO8vfVdcSTud4EIb6jcZ8inLeWucJE0DnxXQBjlQ5zlteuuvooGQy3LIyGxhvlOA==",
+      "version": "14.0.0",
+      "resolved": "https://registry.npmjs.org/copy-webpack-plugin/-/copy-webpack-plugin-14.0.0.tgz",
+      "integrity": "sha512-3JLW90aBGeaTLpM7mYQKpnVdgsUZRExY55giiZgLuX/xTQRUs1dOCwbBnWnvY6Q6rfZoXMNwzOQJCSZPppfqXA==",
      "dev": true,
      "license": "MIT",
      "dependencies": {
-        "fast-glob": "^3.3.2",
        "glob-parent": "^6.0.1",
-        "globby": "^14.0.0",
        "normalize-path": "^3.0.0",
        "schema-utils": "^4.2.0",
-        "serialize-javascript": "^6.0.2"
+        "serialize-javascript": "^7.0.3",
+        "tinyglobby": "^0.2.12"
      },
      "engines": {
-        "node": ">= 18.12.0"
+        "node": ">= 20.9.0"
      },
      "funding": {
        "type": "opencollective",
@@ -8062,36 +8010,6 @@
      "dev": true,
      "license": "Apache-2.0"
    },
-    "node_modules/fast-glob": {
-      "version": "3.3.3",
-      "resolved": "https://registry.npmjs.org/fast-glob/-/fast-glob-3.3.3.tgz",
-      "integrity": "sha512-7MptL8U0cqcFdzIzwOTHoilX9x5BrNqye7Z/LuC7kCMRio1EMSyqRK3BEAUD7sXRq4iT4AzTVuZdhgQ2TCvYLg==",
-      "dev": true,
-      "license": "MIT",
-      "dependencies": {
-        "@nodelib/fs.stat": "^2.0.2",
-        "@nodelib/fs.walk": "^1.2.3",
-        "glob-parent": "^5.1.2",
-        "merge2": "^1.3.0",
-        "micromatch": "^4.0.8"
-      },
-      "engines": {
-        "node": ">=8.6.0"
-      }
-    },
-    "node_modules/fast-glob/node_modules/glob-parent": {
-      "version": "5.1.2",
-      "resolved": "https://registry.npmjs.org/glob-parent/-/glob-parent-5.1.2.tgz",
-      "integrity": "sha512-AOIgSQCepiJYwP3ARnGx+5VnTu2HBYdzbGP45eLw1vr3zB3vZLeyed1sC9hnbcOc9/SrMyM5RPQrkGz4aS9Zow==",
-      "dev": true,
-      "license": "ISC",
-      "dependencies": {
-        "is-glob": "^4.0.1"
-      },
-      "engines": {
-        "node": ">= 6"
-      }
-    },
    "node_modules/fast-json-stable-stringify": {
      "version": "2.1.0",
      "resolved": "https://registry.npmjs.org/fast-json-stable-stringify/-/fast-json-stable-stringify-2.1.0.tgz",
@@ -8180,16 +8098,6 @@
        "node": ">= 4.9.1"
      }
    },
-    "node_modules/fastq": {
-      "version": "1.20.1",
-      "resolved": "https://registry.npmjs.org/fastq/-/fastq-1.20.1.tgz",
-      "integrity": "sha512-GGToxJ/w1x32s/D2EKND7kTil4n8OVk/9mycTc4VDza13lOvpUZTGX3mFSCtV9ksdGBVzvsyAVLM6mHFThxXxw==",
-      "dev": true,
-      "license": "ISC",
-      "dependencies": {
-        "reusify": "^1.0.4"
-      }
-    },
    "node_modules/faye-websocket": {
      "version": "0.11.4",
      "resolved": "https://registry.npmjs.org/faye-websocket/-/faye-websocket-0.11.4.tgz",
@@ -8758,37 +8666,6 @@
        "url": "https://github.com/sponsors/ljharb"
      }
    },
-    "node_modules/globby": {
-      "version": "14.1.0",
-      "resolved": "https://registry.npmjs.org/globby/-/globby-14.1.0.tgz",
-      "integrity": "sha512-0Ia46fDOaT7k4og1PDW4YbodWWr3scS2vAr2lTbsplOt2WkKp0vQbkI9wKis/T5LV/dqPjO3bpS/z6GTJB82LA==",
-      "dev": true,
-      "license": "MIT",
-      "dependencies": {
-        "@sindresorhus/merge-streams": "^2.1.0",
-        "fast-glob": "^3.3.3",
-        "ignore": "^7.0.3",
-        "path-type": "^6.0.0",
-        "slash": "^5.1.0",
-        "unicorn-magic": "^0.3.0"
-      },
-      "engines": {
-        "node": ">=18"
-      },
-      "funding": {
-        "url": "https://github.com/sponsors/sindresorhus"
-      }
-    },
-    "node_modules/globby/node_modules/ignore": {
-      "version": "7.0.5",
-      "resolved": "https://registry.npmjs.org/ignore/-/ignore-7.0.5.tgz",
-      "integrity": "sha512-Hs59xBNfUIunMFgWAbGX5cq6893IbWg4KnrjbYwX3tx0ztorVgTDA6B2sxf8ejHJ4wz8BqGUMYlnzNBer5NvGg==",
-      "dev": true,
-      "license": "MIT",
-      "engines": {
-        "node": ">= 4"
-      }
-    },
    "node_modules/gopd": {
      "version": "1.2.0",
      "resolved": "https://registry.npmjs.org/gopd/-/gopd-1.2.0.tgz",
@@ -8996,23 +8873,6 @@
        "safe-buffer": "~5.1.0"
      }
    },
-    "node_modules/html-entities": {
-      "version": "2.6.0",
-      "resolved": "https://registry.npmjs.org/html-entities/-/html-entities-2.6.0.tgz",
-      "integrity": "sha512-kig+rMn/QOVRvr7c86gQ8lWXq+Hkv6CbAH1hLu+RG338StTpE8Z0b44SDVaqVu7HGKf27frdmUYEs9hTUX/cLQ==",
-      "dev": true,
-      "funding": [
-        {
-          "type": "github",
-          "url": "https://github.com/sponsors/mdevils"
-        },
-        {
-          "type": "patreon",
-          "url": "https://patreon.com/mdevils"
-        }
-      ],
-      "license": "MIT"
-    },
    "node_modules/html-escaper": {
      "version": "3.0.3",
      "resolved": "https://registry.npmjs.org/html-escaper/-/html-escaper-3.0.3.tgz",
@@ -11061,16 +10921,6 @@
      "dev": true,
      "license": "MIT"
    },
-    "node_modules/merge2": {
-      "version": "1.4.1",
-      "resolved": "https://registry.npmjs.org/merge2/-/merge2-1.4.1.tgz",
-      "integrity": "sha512-8q7VEgMJW4J8tcfVPy8g09NcQwZdbwFEqhe/WZkoIzjn/3TGDwtOCYtXGxA3O8tPzpczCCDgv+P2P5y00ZJOOg==",
-      "dev": true,
-      "license": "MIT",
-      "engines": {
-        "node": ">= 8"
-      }
-    },
    "node_modules/methods": {
      "version": "1.1.2",
      "resolved": "https://registry.npmjs.org/methods/-/methods-1.1.2.tgz",
@@ -12655,19 +12505,6 @@
      "dev": true,
      "license": "MIT"
    },
-    "node_modules/path-type": {
-      "version": "6.0.0",
-      "resolved": "https://registry.npmjs.org/path-type/-/path-type-6.0.0.tgz",
-      "integrity": "sha512-Vj7sf++t5pBD637NSfkxpHSMfWaeig5+DKWLhcqIYx6mWQz5hdJTGDVMQiJcw1ZYkhs7AazKDGpRVji1LJCZUQ==",
-      "dev": true,
-      "license": "MIT",
-      "engines": {
-        "node": ">=18"
-      },
-      "funding": {
-        "url": "https://github.com/sponsors/sindresorhus"
-      }
-    },
    "node_modules/pathval": {
      "version": "1.1.1",
      "resolved": "https://registry.npmjs.org/pathval/-/pathval-1.1.1.tgz",
@@ -13002,37 +12839,6 @@
        "url": "https://github.com/sponsors/ljharb"
      }
    },
-    "node_modules/queue-microtask": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/queue-microtask/-/queue-microtask-1.2.3.tgz",
-      "integrity": "sha512-NuaNSa6flKT5JaSYQzJok04JzTL1CA6aGhv5rfLW3PgqA+M2ChpZQnAC8h8i4ZFkBS8X5RqkDBHA7r4hej3K9A==",
-      "dev": true,
-      "funding": [
-        {
-          "type": "github",
-          "url": "https://github.com/sponsors/feross"
-        },
-        {
-          "type": "patreon",
-          "url": "https://www.patreon.com/feross"
-        },
-        {
-          "type": "consulting",
-          "url": "https://feross.org/support"
-        }
-      ],
-      "license": "MIT"
-    },
-    "node_modules/randombytes": {
-      "version": "2.1.0",
-      "resolved": "https://registry.npmjs.org/randombytes/-/randombytes-2.1.0.tgz",
-      "integrity": "sha512-vYl3iOX+4CKUWuxGi9Ukhie6fsqXqS9FE2Zaic4tNFD2N2QQaXOMFbuKK4QmDHC0JO6B1Zp41J0LpT0oR68amQ==",
-      "dev": true,
-      "license": "MIT",
-      "dependencies": {
-        "safe-buffer": "^5.1.0"
-      }
-    },
    "node_modules/range-parser": {
      "version": "1.2.1",
      "resolved": "https://registry.npmjs.org/range-parser/-/range-parser-1.2.1.tgz",
@@ -13519,17 +13325,6 @@
        "node": ">= 4"
      }
    },
-    "node_modules/reusify": {
-      "version": "1.1.0",
-      "resolved": "https://registry.npmjs.org/reusify/-/reusify-1.1.0.tgz",
-      "integrity": "sha512-g6QUff04oZpHs0eG5p83rFLhHeV00ug/Yf9nZM6fLeUrPguBTkTQOdpAWWspMh55TZfVQDPaN3NQJfbVRAxdIw==",
-      "dev": true,
-      "license": "MIT",
-      "engines": {
-        "iojs": ">=1.0.0",
-        "node": ">=0.10.0"
-      }
-    },
    "node_modules/rfdc": {
      "version": "1.4.1",
      "resolved": "https://registry.npmjs.org/rfdc/-/rfdc-1.4.1.tgz",
@@ -13562,30 +13357,6 @@
        "node": ">=0.12.0"
      }
    },
-    "node_modules/run-parallel": {
-      "version": "1.2.0",
-      "resolved": "https://registry.npmjs.org/run-parallel/-/run-parallel-1.2.0.tgz",
-      "integrity": "sha512-5l4VyZR86LZ/lDxZTR6jqL8AFE2S0IFLMP26AbjsLVADxHdhB/c0GUsH+y39UfCi3dzz8OlQuPmnaJOMoDHQBA==",
-      "dev": true,
-      "funding": [
-        {
-          "type": "github",
-          "url": "https://github.com/sponsors/feross"
-        },
-        {
-          "type": "patreon",
-          "url": "https://www.patreon.com/feross"
-        },
-        {
-          "type": "consulting",
-          "url": "https://feross.org/support"
-        }
-      ],
-      "license": "MIT",
-      "dependencies": {
-        "queue-microtask": "^1.2.2"
-      }
-    },
    "node_modules/rxjs": {
      "version": "7.8.2",
      "resolved": "https://registry.npmjs.org/rxjs/-/rxjs-7.8.2.tgz",
@@ -13859,13 +13630,13 @@
      }
    },
    "node_modules/serialize-javascript": {
-      "version": "6.0.2",
-      "resolved": "https://registry.npmjs.org/serialize-javascript/-/serialize-javascript-6.0.2.tgz",
-      "integrity": "sha512-Saa1xPByTTq2gdeFZYLLo+RFE35NHZkAbqZeWNd3BpzppeVisAqpDjcp8dyf6uIvEqJRd46jemmyA4iFIeVk8g==",
+      "version": "7.0.5",
+      "resolved": "https://registry.npmjs.org/serialize-javascript/-/serialize-javascript-7.0.5.tgz",
+      "integrity": "sha512-F4LcB0UqUl1zErq+1nYEEzSHJnIwb3AF2XWB94b+afhrekOUijwooAYqFyRbjYkm2PAKBabx6oYv/xDxNi8IBw==",
      "dev": true,
      "license": "BSD-3-Clause",
-      "dependencies": {
-        "randombytes": "^2.1.0"
+      "engines": {
+        "node": ">=20.0.0"
      }
    },
    "node_modules/serve-index": {
@@ -14277,19 +14048,6 @@
        "simple-concat": "^1.0.0"
      }
    },
-    "node_modules/slash": {
-      "version": "5.1.0",
-      "resolved": "https://registry.npmjs.org/slash/-/slash-5.1.0.tgz",
-      "integrity": "sha512-ZA6oR3T/pEyuqwMgAKT0/hAv8oAXckzbkmR0UkUosQ+Mc4RxGoJkRmwHgHufaenlyAgE1Mxgpdcrf75y6XcnDg==",
-      "dev": true,
-      "license": "MIT",
-      "engines": {
-        "node": ">=14.16"
-      },
-      "funding": {
-        "url": "https://github.com/sponsors/sindresorhus"
-      }
-    },
    "node_modules/sockjs": {
      "version": "0.3.24",
      "resolved": "https://registry.npmjs.org/sockjs/-/sockjs-0.3.24.tgz",
@@ -15328,19 +15086,6 @@
        "node": ">=4"
      }
    },
-    "node_modules/unicorn-magic": {
-      "version": "0.3.0",
-      "resolved": "https://registry.npmjs.org/unicorn-magic/-/unicorn-magic-0.3.0.tgz",
-      "integrity": "sha512-+QBBXBCvifc56fsbuxZQ6Sic3wqqc3WWaqxs58gvJrcOuN83HGTCwz3oS5phzU9LthRNE9VrJCFCLUgHeeFnfA==",
-      "dev": true,
-      "license": "MIT",
-      "engines": {
-        "node": ">=18"
-      },
-      "funding": {
-        "url": "https://github.com/sponsors/sindresorhus"
-      }
-    },
    "node_modules/universalify": {
      "version": "2.0.1",
      "resolved": "https://registry.npmjs.org/universalify/-/universalify-2.0.1.tgz",
@@ -15671,15 +15416,16 @@
      }
    },
    "node_modules/webpack-dev-server": {
-      "version": "5.1.0",
-      "resolved": "https://registry.npmjs.org/webpack-dev-server/-/webpack-dev-server-5.1.0.tgz",
-      "integrity": "sha512-aQpaN81X6tXie1FoOB7xlMfCsN19pSvRAeYUHOdFWOlhpQ/LlbfTqYwwmEDFV0h8GGuqmCmKmT+pxcUV/Nt2gQ==",
+      "version": "5.2.1",
+      "resolved": "https://registry.npmjs.org/webpack-dev-server/-/webpack-dev-server-5.2.1.tgz",
+      "integrity": "sha512-ml/0HIj9NLpVKOMq+SuBPLHcmbG+TGIjXRHsYfZwocUBIqEvws8NnS/V9AFQ5FKP+tgn5adwVwRrTEpGL33QFQ==",
      "dev": true,
      "license": "MIT",
      "dependencies": {
        "@types/bonjour": "^3.5.13",
        "@types/connect-history-api-fallback": "^1.5.4",
        "@types/express": "^4.17.21",
+        "@types/express-serve-static-core": "^4.17.21",
        "@types/serve-index": "^1.9.4",
        "@types/serve-static": "^1.15.5",
        "@types/sockjs": "^0.3.36",
@@ -15690,10 +15436,9 @@
        "colorette": "^2.0.10",
        "compression": "^1.7.4",
        "connect-history-api-fallback": "^2.0.0",
-        "express": "^4.19.2",
+        "express": "^4.21.2",
        "graceful-fs": "^4.2.6",
-        "html-entities": "^2.4.0",
-        "http-proxy-middleware": "^2.0.3",
+        "http-proxy-middleware": "^2.0.7",
        "ipaddr.js": "^2.1.0",
        "launch-editor": "^2.6.1",
        "open": "^10.0.3",
@@ -15728,6 +15473,19 @@
        }
      }
    },
+    "node_modules/webpack-dev-server/node_modules/@types/express-serve-static-core": {
+      "version": "4.19.8",
+      "resolved": "https://registry.npmjs.org/@types/express-serve-static-core/-/express-serve-static-core-4.19.8.tgz",
+      "integrity": "sha512-02S5fmqeoKzVZCHPZid4b8JH2eM5HzQLZWN2FohQEy/0eXTq8VXZfSN6Pcr3F6N9R/vNrj7cpgbhjie6m/1tCA==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "@types/node": "*",
+        "@types/qs": "*",
+        "@types/range-parser": "*",
+        "@types/send": "*"
+      }
+    },
    "node_modules/webpack-dev-server/node_modules/define-lazy-prop": {
      "version": "3.0.0",
      "resolved": "https://registry.npmjs.org/define-lazy-prop/-/define-lazy-prop-3.0.0.tgz",
@@ -15771,9 +15529,9 @@
      }
    },
    "node_modules/webpack-dev-server/node_modules/ws": {
-      "version": "8.19.0",
-      "resolved": "https://registry.npmjs.org/ws/-/ws-8.19.0.tgz",
-      "integrity": "sha512-blAT2mjOEIi0ZzruJfIhb3nps74PRWTCz1IjglWEEpQl5XS/UNama6u2/rjFkDDouqr4L67ry+1aGIALViWjDg==",
+      "version": "8.20.0",
+      "resolved": "https://registry.npmjs.org/ws/-/ws-8.20.0.tgz",
+      "integrity": "sha512-sAt8BhgNbzCtgGbt2OxmpuryO63ZoDk/sqaB/znQm94T4fCEsy/yV+7CdC1kJhOU9lboAEU7R3kquuycDoibVA==",
      "dev": true,
      "license": "MIT",
      "engines": {
@@ -36,7 +36,7 @@
    "@types/office-runtime": "^1.0.35",
    "acorn": "^8.11.3",
    "babel-loader": "^9.1.3",
-    "copy-webpack-plugin": "^12.0.2",
+    "copy-webpack-plugin": "^14.0.0",
    "eslint-plugin-office-addins": "^4.0.3",
    "file-loader": "^6.2.0",
    "html-loader": "^5.0.0",
@@ -53,7 +53,7 @@
    "source-map-loader": "^5.0.0",
    "webpack": "^5.95.0",
    "webpack-cli": "^5.1.4",
-    "webpack-dev-server": "5.1.0"
+    "webpack-dev-server": "5.2.1"
  },
  "prettier": "office-addin-prettier-config",
  "browserslist": [
@@ -1,17 +0,0 @@
-{
-  "name": "thunderbird",
-  "version": "1.0.0",
-  "main": "index.js",
-  "private": true,
-  "scripts": {
-    "dev": "web-ext run",
-    "lint": "web-ext lint --source-dir=./src"
-  },
-  "keywords": [],
-  "author": "",
-  "license": "ISC",
-  "description": "",
-  "devDependencies": {
-    "web-ext": "^10.1.0"
-  }
-}
@@ -1,48 +0,0 @@
-import { buildMeetingMessage } from "./lib/meeting-message.js";
-
-console.log("[meeting-link] background loaded at", new Date().toISOString());
-
-(async () => {
-  try {
-    console.log("[meeting-link] browser.calendar exists?", !!browser.calendar);
-    if (browser.calendar) {
-      console.log("[meeting-link] calendar namespace keys:",
-        Object.keys(browser.calendar));
-    }
-
-    // Try the most basic read call — list configured calendars.
-    // The exact method name varies between experiment versions; we'll
-    // try the common one first.
-    if (browser.calendar.calendars?.query) {
-      const calendars = await browser.calendar.calendars.query({});
-      console.log("[meeting-link] calendars found:", calendars.length, calendars);
-    } else {
-      console.warn("[meeting-link] calendars.query not present — namespace shape:",
-        browser.calendar);
-    }
-
-  } catch (err) {
-    console.error("[meeting-link] calendar probe failed:", err);
-  }
-})();
-
-// Hardcoded for Spike 1. Spike 2 replaces this with a fetch() to your API.
-const STUB_MEETING_DATA = {
-  url: "https://meet.example.com/m/abc-123-xyz",
-  telephony: {
-    phone_number: "+33123456789",
-    pin_code: "1234567890",
-  },
-};
-
-browser.runtime.onMessage.addListener(async (msg) => {
-  if (msg?.type === "GET_MEETING_MESSAGE") {
-    try {
-      const built = buildMeetingMessage(STUB_MEETING_DATA);
-      return { ok: true, ...built };
-    } catch (err) {
-      console.error("[meeting-link] build failed", err);
-      return { ok: false, error: String(err.message || err) };
-    }
-  }
-});
@@ -1,41 +0,0 @@
-{
-  "manifest_version": 3,
-  "name": "Meeting Link (dev)",
-  "version": "0.1.0",
-  "description": "Spike 0 — verifying the dev loop",
-  "browser_specific_settings": {
-    "gecko": {
-      "id": "meeting-link-dev@yourcompany.example",
-      "strict_min_version": "128.0"
-    }
-  },
-  "background": {
-    "scripts": ["background.js"],
-    "type": "module"
-  },
-  "compose_action": {
-    "default_title": "Insert meeting link",
-    "default_popup": "popup/popup.html",
-    "default_icon": "icons/icon-32.png"
-  },
-  "permissions": ["compose"],
-  "experiment_apis": {
-    "calendar_provider": {
-      "schema": "experiments/calendar/schema/calendar-provider.json",
-      "parent": {
-        "scopes": ["addon_parent"],
-        "paths": [["calendar", "provider"]],
-        "script": "experiments/calendar/parent/ext-calendar-provider.js",
-        "events": ["startup"]
-      }
-    },
-    "calendar_calendars": {
-      "schema": "experiments/calendar/schema/calendar-calendars.json",
-      "parent": {
-        "scopes": ["addon_parent"],
-        "paths": [["calendar", "calendars"]],
-        "script": "experiments/calendar/parent/ext-calendar-calendars.js"
-      }
-    }
-  }
-}
@@ -1,18 +0,0 @@
-<!doctype html>
-<html>
-    <head>
-        <meta charset="utf-8">
-        <title>Meeting link</title>
-        <style>
-            body { font: 13px system-ui; padding: 12px; min-width: 240px; margin: 0; }
-            button { font: inherit; padding: 6px 12px; cursor: pointer; }
-            #status { margin-top: 8px; color: #555; min-height: 1.2em; white-space: pre-wrap; }
-            #status.error { color: #b00020; }
-        </style>
-    </head>
-    <body>
-        <button id="insert">Insert meeting link</button>
-        <div id="status"></div>
-        <script src="popup.js"></script>
-    </body>
-</html>
@@ -1,65 +0,0 @@
-const insertBtn = document.getElementById("insert");
-const statusEl = document.getElementById("status");
-
-function setStatus(text, isError = false) {
-  statusEl.textContent = text;
-  statusEl.classList.toggle("error", isError);
-}
-
-insertBtn.addEventListener("click", async () => {
-  insertBtn.disabled = true;
-  setStatus("Generating link…");
-
-  try {
-    // 1. Find the compose tab this popup belongs to.
-    //    A compose_action popup is anchored to a compose window, so the
-    //    "active tab in the current window" is the compose tab itself.
-    const [composeTab] = await browser.tabs.query({
-      active: true,
-      currentWindow: true,
-    });
-    if (!composeTab) throw new Error("No compose tab found");
-
-    // 2. Read the current compose state — we need to know if we're
-    //    in HTML mode or plain-text mode, and we need the existing body
-    //    so we can append rather than overwrite.
-    const details = await browser.compose.getComposeDetails(composeTab.id);
-
-    // 3. Ask the background to produce the meeting message.
-    const reply = await browser.runtime.sendMessage({
-      type: "GET_MEETING_MESSAGE",
-    });
-    if (!reply?.ok) throw new Error(reply?.error || "Background error");
-
-    // 4. Append to the existing body in the right format.
-    if (details.isPlainText) {
-      const newBody = (details.plainTextBody || "") + "\n\n" + reply.text;
-      await browser.compose.setComposeDetails(composeTab.id, {
-        plainTextBody: newBody,
-      });
-    } else {
-      const newBody = appendHtmlBeforeBodyEnd(details.body || "", reply.html);
-      await browser.compose.setComposeDetails(composeTab.id, {
-        body: newBody,
-      });
-    }
-
-    setStatus("Inserted ✓");
-    setTimeout(() => window.close(), 600);
-  } catch (err) {
-    console.error("[meeting-link] popup insert failed", err);
-    setStatus("Failed: " + (err.message || err), true);
-    insertBtn.disabled = false;
-  }
-});
-
-/**
- * Append HTML right before </body>, or fall back to concatenation if no
- * </body> tag is present (Thunderbird's compose body is usually a full
- * HTML document, but be defensive).
- */
-function appendHtmlBeforeBodyEnd(currentHtml, fragment) {
-  const idx = currentHtml.toLowerCase().lastIndexOf("</body>");
-  if (idx === -1) return currentHtml + fragment;
-  return currentHtml.slice(0, idx) + fragment + currentHtml.slice(idx);
-}
@@ -1,30 +0,0 @@
-// web-ext-config.cjs
-const os = require("os");
-const path = require("path");
-
-function thunderbirdBinary() {
-  if (process.env.WEB_EXT_FIREFOX) return process.env.WEB_EXT_FIREFOX;
-  switch (process.platform) {
-    case "darwin":
-      return "/Applications/Thunderbird Beta.app/Contents/MacOS/thunderbird";
-    case "linux":
-      return "/usr/bin/thunderbird"; // adjust to your install
-    case "win32":
-      return "C:\\Program Files\\Mozilla Thunderbird\\thunderbird.exe";
-    default:
-      throw new Error("Unsupported platform: " + process.platform);
-  }
-}
-
-module.exports = {
-  sourceDir: "./src",
-  artifactsDir: "./web-ext-artifacts",
-  run: {
-    firefox: thunderbirdBinary(),
-    firefoxProfile: path.join(os.homedir(), ".thunderbird-dev-profile"),
-    profileCreateIfMissing: true,
-    keepProfileChanges: true,
-    browserConsole: true,
-  },
-  ignoreFiles: ["package-lock.json", "web-ext-config.cjs", "*.md"],
-};
@@ -0,0 +1,38 @@
+# Python
+__pycache__
+*.pyc
+**/__pycache__
+**/*.pyc
+venv
+**/.venv
+
+# System-specific files
+.DS_Store
+**/.DS_Store
+
+# Docker
+compose.*
+env.d
+
+# Docs
+docs
+*.md
+*.log
+
+# Development/test cache & configurations
+data
+.cache
+.circleci
+.git
+.iml
+db.sqlite3
+.pylint.d
+
+**/.idea
+**/.vscode
+**/.pytest_cache
+**/.mypy_cache
+**/.ruff_cache
+
+# Env
+.env
@@ -6,31 +6,61 @@ RUN apt-get update && apt-get install -y \
    libgobject-2.0-0 \
    && rm -rf /var/lib/apt/lists/*

+
+# ---- Builder image ----
 FROM base AS builder

-WORKDIR /builder
+ENV UV_COMPILE_BYTECODE=1 \
+    UV_LINK_MODE=copy \
+    UV_PYTHON_DOWNLOADS=0

-COPY pyproject.toml .
-
-RUN mkdir /install && \
-  pip install --prefix=/install .
-
-FROM base AS development
+# Install uv
+COPY --from=ghcr.io/astral-sh/uv:0.10.9 /uv /uvx /bin/

 WORKDIR /app

-COPY pyproject.toml .
-RUN pip install --no-cache-dir ".[dev]"
+# Install production dependencies without the project itself (cacheable layer)
+RUN --mount=type=cache,target=/root/.cache/uv \
+    --mount=type=bind,source=uv.lock,target=uv.lock \
+    --mount=type=bind,source=pyproject.toml,target=pyproject.toml \
+    uv sync --locked --no-install-project --no-dev

-COPY . .
+# Install the project
+COPY . /app
+RUN --mount=type=cache,target=/root/.cache/uv \
+    uv sync --locked --no-dev

-CMD ["python", "metadata_collector.py", "dev"]

+# ---- Development image ----
+FROM base AS development
+
+ENV UV_COMPILE_BYTECODE=1 \
+    UV_LINK_MODE=copy \
+    UV_PYTHON_DOWNLOADS=0
+
+COPY --from=ghcr.io/astral-sh/uv:0.10.9 /uv /uvx /bin/
+
+WORKDIR /app
+
+COPY . /app
+
+RUN --mount=type=cache,target=/root/.cache/uv \
+    uv sync --locked --all-extras
+
+ENV PATH="/app/.venv/bin:$PATH"
+
+CMD ["python", "multi_user_transcriber.py", "dev"]
+
+
+# ---- Production image ----
 FROM base AS production

 WORKDIR /app

-COPY --from=builder /install /usr/local
+# Copy the pre-built virtualenv and application source
+COPY --from=builder /app /app
+
+ENV PATH="/app/.venv/bin:$PATH"

 # Remove pip to reduce attack surface in production
 RUN pip uninstall -y pip
@@ -39,6 +69,4 @@ RUN pip uninstall -y pip
 ARG DOCKER_USER
 USER ${DOCKER_USER}

-COPY ./*.py /app/
-
 CMD ["python", "multi_user_transcriber.py", "start"]
@@ -177,7 +177,7 @@ class MetadataCollector:

    def save(self):
        """Serialize collected events and upload as JSON to S3."""
-        logger.info("Persisting metadata…")
+        logger.info("Persisting metadata...")

        participants = []
        for k, v in self.participants.items():
@@ -372,7 +372,7 @@ async def entrypoint(ctx: JobContext):
    await ctx.connect(auto_subscribe=AutoSubscribe.AUDIO_ONLY)

    async def cleanup():
-        logger.info("Shutting down metadata collector…")
+        logger.info("Shutting down metadata collector...")
        await metadata_collector.aclose()

    ctx.add_shutdown_callback(cleanup)
@@ -18,12 +18,8 @@ dev = [
    "ruff==0.15.6",
 ]

-[tool.setuptools]
-py-modules = ["multi_user_transcriber", "metadata_collector", "exceptions"]
-
-[build-system]
-requires = ["setuptools>=61.0"]
-build-backend = "setuptools.build_meta"
+[tool.uv]
+package = false

 [tool.ruff]
 target-version = "py313"
@@ -14,6 +14,7 @@ from django.utils.translation import gettext_lazy as _

 from django_pydantic_field.rest_framework import SchemaField
 from pydantic import BaseModel, Field
+from pydantic import ValidationError as PydanticValidationError
 from rest_framework import serializers
 from rest_framework.exceptions import PermissionDenied
 from timezone_field.rest_framework import TimeZoneSerializerField
@@ -131,6 +132,16 @@ class RoomSerializer(serializers.ModelSerializer):
        fields = ["id", "name", "slug", "configuration", "access_level", "pin_code"]
        read_only_fields = ["id", "slug", "pin_code"]

+    def validate_configuration(self, value):
+        """Validate room configuration against the RoomConfiguration schema."""
+        if value is None or value == {}:
+            return value
+        try:
+            RoomConfiguration.model_validate(value)
+        except PydanticValidationError as e:
+            raise serializers.ValidationError(e.errors()) from e
+        return value
+
    def to_representation(self, instance):
        """
        Add users only for administrator users.
@@ -306,6 +317,22 @@ class MuteParticipantSerializer(BaseParticipantsManagementSerializer):
    )


+RoomConfigurationTrackSource = Literal[
+    "camera", "microphone", "screen_share", "screen_share_audio"
+]
+
+
+class RoomConfiguration(BaseModel):
+    """Validate room configuration structure.
+
+    Unknown fields are rejected.
+    """
+
+    can_publish_sources: list[RoomConfigurationTrackSource] | None = None
+
+    model_config = {"extra": "forbid"}
+
+
 TrackSource = Literal["SCREEN_SHARE", "SCREEN_SHARE_AUDIO", "CAMERA", "MICROPHONE"]


@@ -6,7 +6,9 @@ from logging import getLogger
 from urllib.parse import unquote, urlparse

 from django.conf import settings
+from django.core.exceptions import ValidationError as DjangoValidationError
 from django.core.files.storage import default_storage
+from django.db import IntegrityError, transaction
 from django.db.models import Q
 from django.http import Http404
 from django.shortcuts import get_object_or_404
@@ -320,16 +322,27 @@ class RoomViewSet(
        options = serializer.validated_data.get("options")
        room = self.get_object()

-        # May raise exception if an active or initiated recording already exist for the room
-        recording = models.Recording.objects.create(
-            room=room,
-            mode=mode,
-            options=options.model_dump(exclude_none=True) if options else {},
-        )
+        try:
+            with transaction.atomic():
+                recording = models.Recording.objects.create(
+                    room=room,
+                    mode=mode,
+                    options=options.model_dump(exclude_none=True) if options else {},
+                )
+                models.RecordingAccess.objects.create(
+                    user=self.request.user,
+                    role=models.RoleChoices.OWNER,
+                    recording=recording,
+                )

-        models.RecordingAccess.objects.create(
-            user=self.request.user, role=models.RoleChoices.OWNER, recording=recording
-        )
+        except (DjangoValidationError, IntegrityError):
+            # DjangoValidationError covers the Python-level check (full_clean);
+            # IntegrityError covers the race where two concurrent requests both
+            # pass that check and the DB-level UNIQUE constraint catches the loser.
+            return drf_response.Response(
+                {"error": f"A recording is already in progress for room {room.slug}"},
+                status=drf_status.HTTP_409_CONFLICT,
+            )

        worker_service = get_worker_service(mode=recording.mode)
        worker_manager = WorkerServiceMediator(worker_service=worker_service)
@@ -337,9 +350,12 @@ class RoomViewSet(
        try:
            worker_manager.start(recording)
        except RecordingStartError:
+            models.Recording.objects.filter(pk=recording.pk).update(
+                status=models.RecordingStatusChoices.FAILED_TO_START
+            )
            return drf_response.Response(
                {"error": f"Recording failed to start for room {room.slug}"},
-                status=drf_status.HTTP_500_INTERNAL_SERVER_ERROR,
+                status=drf_status.HTTP_502_BAD_GATEWAY,
            )

        if settings.METADATA_COLLECTOR_ENABLED and (
@@ -347,6 +363,7 @@ class RoomViewSet(
        ):
            try:
                MetadataCollectorService().start(recording)
+                logger.debug("Started MetadataCollectorService")
            except MetadataCollectorException:
                logger.warning("Failed to start MetadataCollectorService")

@@ -1,7 +1,9 @@
 """Service to notify external services when a new recording is ready."""

+import asyncio
 import logging
 import smtplib
+from datetime import datetime, timezone

 from django.conf import settings
 from django.core.mail import send_mail
@@ -9,9 +11,12 @@ from django.template.loader import render_to_string
 from django.utils.translation import get_language, override
 from django.utils.translation import gettext_lazy as _

+import aiohttp
 import requests
+from asgiref.sync import async_to_sync
+from livekit import api as livekit_api

-from core import models
+from core import models, utils

 logger = logging.getLogger(__name__)

@@ -131,7 +136,50 @@ class NotificationService:
        return not has_failures

    @staticmethod
-    def _notify_summary_service(recording):
+    async def _get_recording_timestamps(worker_id):
+        """Fetch FileInfo.started_at and ended_at from LiveKit's egress API.
+
+        FileInfo.started_at is more accurate than EgressInfo.started_at because
+        it reflects when file recording actually began. The started_at value exposed
+        in the manifest file, as well as in the EgressInfo returned by the API,
+        corresponds to when the egress service received the request, not the moment
+        the egress worker effectively joined the room.
+
+        Returns:
+            Tuple of (started_at, ended_at) datetimes, either may be None.
+        """
+
+        if not worker_id:
+            return None, None
+
+        custom_configuration = {
+            **settings.LIVEKIT_CONFIGURATION,
+            "timeout": aiohttp.ClientTimeout(total=10),
+        }
+        lkapi = utils.create_livekit_client(custom_configuration=custom_configuration)
+        try:
+            egress_list = await lkapi.egress.list_egress(
+                livekit_api.ListEgressRequest(egress_id=worker_id)  # pylint: disable=no-member
+            )
+        except (livekit_api.TwirpError, OSError, asyncio.TimeoutError):
+            logger.exception("Could not fetch egress info for worker %s", worker_id)
+            return None, None
+        finally:
+            await lkapi.aclose()
+
+        if not egress_list.items or not egress_list.items[0].file_results:
+            logger.debug("No file_results for worker %s", worker_id)
+            return None, None
+
+        file_result = egress_list.items[0].file_results[0]
+
+        def _ns_to_utc(ns):
+            return datetime.fromtimestamp(ns / 1e9, tz=timezone.utc) if ns else None
+
+        return _ns_to_utc(file_result.started_at), _ns_to_utc(file_result.ended_at)
+
+    @staticmethod
+    def _notify_summary_service(recording: models.Recording):
        """Notify summary service about a new recording."""

        if (
@@ -150,24 +198,35 @@ class NotificationService:
            .first()
        )

+        if settings.METADATA_COLLECTOR_ENABLED and recording.options.get(
+            "collect_metadata", False
+        ):
+            output_folder = settings.METADATA_COLLECTOR_OUTPUT_FOLDER
+            metadata_filename = f"{output_folder}/{recording.id}-metadata.json"
+        else:
+            metadata_filename = None
+
        if not owner_access:
            logger.error("No owner found for recording %s", recording.id)
            return False
+
+        started_at, ended_at = async_to_sync(
+            NotificationService._get_recording_timestamps
+        )(recording.worker_id)
+
        payload = {
            "owner_id": str(owner_access.user.id),
-            "filename": recording.key,
+            "recording_filename": recording.key,
+            "metadata_filename": metadata_filename,
            "email": owner_access.user.email,
            "sub": owner_access.user.sub,
            "room": recording.room.name,
            "language": recording.options.get("language"),
-            "recording_date": recording.created_at.astimezone(
-                owner_access.user.timezone
-            ).strftime("%Y-%m-%d"),
-            "recording_time": recording.created_at.astimezone(
-                owner_access.user.timezone
-            ).strftime("%H:%M"),
+            "owner_timezone": str(owner_access.user.timezone),
            "download_link": f"{get_recording_download_base_url()}/{recording.id}",
            "context_language": owner_access.user.language,
+            "recording_start_at": (started_at.isoformat() if started_at else None),
+            "recording_end_at": (ended_at.isoformat() if ended_at else None),
        }

        headers = {
@@ -1,5 +1,7 @@
 """Factory, configurations and Protocol to create worker services"""

+# pylint: disable=no-member
+
 import logging
 from dataclasses import dataclass
 from functools import lru_cache
@@ -8,8 +10,17 @@ from typing import Any, ClassVar, Dict, Optional, Protocol, Type
 from django.conf import settings
 from django.utils.module_loading import import_string

+from livekit import api as livekit_api
+
 logger = logging.getLogger(__name__)

+# Codec / frequency constants matching LiveKit's H264_720P_30 preset.
+# Kept fixed because changing them would shift the goal-post away from the
+# "safe drop-in replacement for the default preset" contract of this feature.
+_RECORDING_VIDEO_CODEC = livekit_api.VideoCodec.H264_MAIN
+_RECORDING_AUDIO_CODEC = livekit_api.AudioCodec.AAC
+_RECORDING_AUDIO_FREQUENCY_HZ = 48000
+

@dataclass(frozen=True)
 class WorkerServiceConfig:
@@ -18,6 +29,7 @@ class WorkerServiceConfig:
    output_folder: str
    server_configurations: Dict[str, Any]
    bucket_args: Optional[dict]
+    encoding_options: Optional[Dict[str, Any]] = None

    @classmethod
    @lru_cache
@@ -25,6 +37,24 @@ class WorkerServiceConfig:
        """Load configuration from Django settings with caching for efficiency."""

        logger.debug("Loading WorkerServiceConfig from settings.")
+
+        encoding_options: Optional[Dict[str, Any]] = None
+        if settings.RECORDING_ENCODING_ENABLED:
+            # Single source of truth for the EncodingOptions kwargs:
+            # operator-tunable values live in Django settings, codec / frequency
+            # are pinned constants. The services layer only unpacks this dict.
+            encoding_options = {
+                "width": settings.RECORDING_ENCODING_WIDTH,
+                "height": settings.RECORDING_ENCODING_HEIGHT,
+                "framerate": settings.RECORDING_ENCODING_FRAMERATE,
+                "video_bitrate": settings.RECORDING_ENCODING_VIDEO_BITRATE_KBPS,
+                "audio_bitrate": settings.RECORDING_ENCODING_AUDIO_BITRATE_KBPS,
+                "key_frame_interval": settings.RECORDING_ENCODING_KEY_FRAME_INTERVAL_S,
+                "video_codec": _RECORDING_VIDEO_CODEC,
+                "audio_codec": _RECORDING_AUDIO_CODEC,
+                "audio_frequency": _RECORDING_AUDIO_FREQUENCY_HZ,
+            }
+
        return cls(
            output_folder=settings.RECORDING_OUTPUT_FOLDER,
            server_configurations=settings.LIVEKIT_CONFIGURATION,
@@ -36,6 +66,7 @@ class WorkerServiceConfig:
                "bucket": settings.AWS_STORAGE_BUCKET_NAME,
                "force_path_style": True,
            },
+            encoding_options=encoding_options,
        )


@@ -83,6 +83,22 @@ class BaseEgressService:
        """
        raise NotImplementedError("Subclass must implement this method.")

+    def _build_encoding_options(self):
+        """Build a LiveKit EncodingOptions from the service config, or None.
+
+        When None is returned, the caller should omit the `advanced` field so
+        LiveKit Egress falls back to its built-in preset (H264_720P_30).
+
+        The full EncodingOptions kwargs (operator-tunable values + pinned
+        codec / frequency constants) are assembled in `WorkerServiceConfig`,
+        so this method is a thin protobuf adapter.
+        """
+        opts = self._config.encoding_options
+        if not opts:
+            return None
+
+        return livekit_api.EncodingOptions(**opts)
+

 class VideoCompositeEgressService(BaseEgressService):
    """Record multiple participant video and audio tracks into a single output '.mp4' file."""
@@ -104,9 +120,17 @@ class VideoCompositeEgressService(BaseEgressService):
            s3=self._s3,
        )

-        request = livekit_api.RoomCompositeEgressRequest(
-            room_name=room_name, file_outputs=[file_output], layout="speaker-light"
-        )
+        request_kwargs = {
+            "room_name": room_name,
+            "file_outputs": [file_output],
+            "layout": "speaker-light",
+        }
+
+        advanced = self._build_encoding_options()
+        if advanced is not None:
+            request_kwargs["advanced"] = advanced
+
+        request = livekit_api.RoomCompositeEgressRequest(**request_kwargs)

        response = self._handle_request(request, "start_room_composite_egress")

@@ -123,7 +123,7 @@ class LobbyService:

    def request_entry(
        self,
-        room,
+        room: models.Room,
        request,
        username: str,
    ) -> Tuple[LobbyParticipant, Optional[Dict]]:
@@ -117,7 +117,7 @@ def test_api_files_create_file_authenticated_success():
    policy_parsed = urlparse(policy)

    assert policy_parsed.scheme == "http"
-    assert policy_parsed.netloc == "localhost:9000"
+    assert policy_parsed.netloc in ["minio:9000", "localhost:9000"]
    assert policy_parsed.path == f"/meet-media-storage/files/{file.id!s}.png"

    query_params = parse_qs(policy_parsed.query)
@@ -2,7 +2,7 @@
 Test worker service factories.
 """

-# pylint: disable=protected-access,redefined-outer-name,unused-argument
+# pylint: disable=protected-access,redefined-outer-name,unused-argument,no-member

 from dataclasses import FrozenInstanceError
 from unittest.mock import Mock
@@ -10,6 +10,9 @@ from unittest.mock import Mock
 from django.test import override_settings

 import pytest
+from livekit import (
+    api as livekit_api_codec,
+)

 from core.recording.worker.factories import (
    WorkerService,
@@ -63,6 +66,8 @@ def test_config_initialization(default_config):
        "bucket": "test-bucket",
        "force_path_style": True,
    }
+    # Encoding override is opt-in; disabled by default.
+    assert default_config.encoding_options is None


 def test_config_immutability(default_config):
@@ -71,6 +76,45 @@ def test_config_immutability(default_config):
        default_config.output_folder = "new/path"


+@override_settings(
+    RECORDING_OUTPUT_FOLDER="/test/output",
+    LIVEKIT_CONFIGURATION={"server": "test.example.com"},
+    AWS_S3_ENDPOINT_URL="https://s3.test.com",
+    AWS_S3_ACCESS_KEY_ID="test_key",
+    AWS_S3_SECRET_ACCESS_KEY="test_secret",
+    AWS_S3_REGION_NAME="test-region",
+    AWS_STORAGE_BUCKET_NAME="test-bucket",
+    RECORDING_ENCODING_ENABLED=True,
+    RECORDING_ENCODING_WIDTH=1280,
+    RECORDING_ENCODING_HEIGHT=720,
+    RECORDING_ENCODING_FRAMERATE=15,
+    RECORDING_ENCODING_VIDEO_BITRATE_KBPS=600,
+    RECORDING_ENCODING_AUDIO_BITRATE_KBPS=64,
+    RECORDING_ENCODING_KEY_FRAME_INTERVAL_S=10.0,
+)
+def test_config_encoding_options_enabled():
+    """When RECORDING_ENCODING_ENABLED is True, encoding options are populated.
+
+    The dict mixes operator-tunable values from settings with pinned codec /
+    frequency constants, so the services layer can simply unpack it.
+    """
+
+    WorkerServiceConfig.from_settings.cache_clear()
+    config = WorkerServiceConfig.from_settings()
+
+    assert config.encoding_options == {
+        "width": 1280,
+        "height": 720,
+        "framerate": 15,
+        "video_bitrate": 600,
+        "audio_bitrate": 64,
+        "key_frame_interval": 10.0,
+        "video_codec": livekit_api_codec.VideoCodec.H264_MAIN,
+        "audio_codec": livekit_api_codec.AudioCodec.AAC,
+        "audio_frequency": 48000,
+    }
+
+
@override_settings(
    RECORDING_OUTPUT_FOLDER="/test/output",
    LIVEKIT_CONFIGURATION={"server": "test.example.com"},
@@ -39,6 +39,31 @@ def config():
    )


+@pytest.fixture
+def config_with_encoding(config):
+    """Fixture for a config carrying custom encoding options.
+
+    Mirrors the dict shape produced by `WorkerServiceConfig.from_settings()`
+    (operator-tunable values + pinned codec / frequency constants).
+    """
+    return WorkerServiceConfig(
+        output_folder=config.output_folder,
+        server_configurations=config.server_configurations,
+        bucket_args=config.bucket_args,
+        encoding_options={
+            "width": 1280,
+            "height": 720,
+            "framerate": 15,
+            "video_bitrate": 600,
+            "audio_bitrate": 64,
+            "key_frame_interval": 10.0,
+            "video_codec": livekit_api.VideoCodec.H264_MAIN,
+            "audio_codec": livekit_api.AudioCodec.AAC,
+            "audio_frequency": 48000,
+        },
+    )
+
+
@pytest.fixture
 def mock_s3_upload():
    """Fixture for mocked S3Upload"""
@@ -224,6 +249,41 @@ def test_video_composite_egress_start_missing_egress_id(video_service):
    assert "Egress ID not found" in str(exc_info.value)


+def test_video_composite_egress_start_without_encoding_options(video_service):
+    """When no encoding options are configured, no `advanced` field is set.
+
+    LiveKit then falls back to its built-in preset (H264_720P_30).
+    """
+    video_service._handle_request.return_value = Mock(egress_id="eg-1")
+    video_service.start("test-room", "rec-1")
+
+    request = video_service._handle_request.call_args[0][0]
+    # Proto oneof `options` must be unset when no advanced encoding is provided.
+    assert request.WhichOneof("options") is None
+
+
+def test_video_composite_egress_start_with_encoding_options(config_with_encoding):
+    """Custom encoding options are forwarded as `advanced` EncodingOptions."""
+    service = VideoCompositeEgressService(config_with_encoding)
+    service._handle_request = Mock(return_value=Mock(egress_id="eg-2"))
+
+    service.start("test-room", "rec-2")
+
+    request = service._handle_request.call_args[0][0]
+    assert request.WhichOneof("options") == "advanced"
+
+    advanced = request.advanced
+    assert advanced.width == 1280
+    assert advanced.height == 720
+    assert advanced.framerate == 15
+    assert advanced.video_bitrate == 600
+    assert advanced.audio_bitrate == 64
+    assert advanced.key_frame_interval == pytest.approx(10.0)
+    assert advanced.video_codec == livekit_api.VideoCodec.H264_MAIN
+    assert advanced.audio_codec == livekit_api.AudioCodec.AAC
+    assert advanced.audio_frequency == 48000
+
+
 def test_audio_composite_egress_hrid(audio_service):
    """Test HRID is correct"""
    assert audio_service.hrid == "audio-recording-composite-livekit-egress"
@@ -232,7 +232,7 @@ def test_api_rooms_retrieve_authenticated_public(mock_token):
    """
    room = RoomFactory(
        access_level=RoomAccessLevel.PUBLIC,
-        configuration={"can_publish_sources": ["mock-source"]},
+        configuration={"can_publish_sources": ["camera"]},
    )

    user = UserFactory()
@@ -264,7 +264,7 @@ def test_api_rooms_retrieve_authenticated_public(mock_token):
        user=user,
        username=None,
        color=None,
-        sources=["mock-source"],
+        sources=["camera"],
        is_admin_or_owner=False,
        participant_id=None,
    )
@@ -363,7 +363,7 @@ def test_api_rooms_retrieve_members(mock_token, django_assert_num_queries, setti
    other_user = UserFactory()

    room = RoomFactory(
-        configuration={"can_publish_sources": ["mock-source"]},
+        configuration={"can_publish_sources": ["camera"]},
    )
    UserResourceAccessFactory(resource=room, user=user, role="member")
    UserResourceAccessFactory(resource=room, user=other_user, role="member")
@@ -401,7 +401,7 @@ def test_api_rooms_retrieve_members(mock_token, django_assert_num_queries, setti
        user=user,
        username=None,
        color=None,
-        sources=["mock-source"],
+        sources=["camera"],
        is_admin_or_owner=False,
        participant_id=None,
    )
@@ -140,16 +140,18 @@ def test_start_recording_worker_error(

    mock_worker_service_factory.assert_called_once_with(mode="screen_recording")

-    assert response.status_code == 500
+    assert response.status_code == 502
    assert response.json() == {
        "error": f"Recording failed to start for room {room.slug}"
    }

-    # Recording object should be created even if worker fails
+    # Recording object should be created even if worker fails, and moved out
+    # of the unique-constraint window so the room is not locked.
    assert Recording.objects.count() == 1
    recording = Recording.objects.first()
    assert recording.room == room
    assert recording.mode == "screen_recording"
+    assert recording.status == "failed_to_start"

    # Verify recording access details
    assert recording.accesses.count() == 1
@@ -158,6 +160,72 @@ def test_start_recording_worker_error(
    assert access.role == "owner"


+@pytest.mark.parametrize(
+    "status",
+    ["active", "initiated"],
+)
+def test_start_recording_conflict_when_already_in_progress(
+    status, mock_worker_service_factory, mock_worker_manager, settings
+):
+    """Should return 409 when a second start is attempted while a recording is already active."""
+    settings.RECORDING_ENABLE = True
+
+    room = RoomFactory()
+    user = UserFactory()
+    room.accesses.create(user=user, role="owner")
+
+    # Pre-existing active recording for the same room.
+    Recording.objects.create(room=room, mode="screen_recording", status="active")
+
+    client = APIClient()
+    client.force_login(user)
+
+    response = client.post(
+        f"/api/v1.0/rooms/{room.id}/start-recording/",
+        {"mode": "screen_recording"},
+    )
+
+    assert response.status_code == 409
+    assert response.json() == {
+        "error": f"A recording is already in progress for room {room.slug}"
+    }
+    # No new recording row, no access row leaked from the rolled-back transaction.
+    assert Recording.objects.count() == 1
+    assert Recording.objects.first().accesses.count() == 0
+    mock_worker_manager.start.assert_not_called()
+
+
+def test_start_recording_after_worker_failure_unblocks_room(
+    mock_worker_service_factory, mock_worker_manager, settings
+):
+    """Should allow a new recording when the previous recording failed."""
+    settings.RECORDING_ENABLE = True
+
+    room = RoomFactory()
+    user = UserFactory()
+    room.accesses.create(user=user, role="owner")
+
+    mock_worker_manager.start = mock.Mock(
+        side_effect=[RecordingStartError("boom"), None]
+    )
+
+    client = APIClient()
+    client.force_login(user)
+
+    first = client.post(
+        f"/api/v1.0/rooms/{room.id}/start-recording/",
+        {"mode": "screen_recording"},
+    )
+    assert first.status_code == 502
+
+    second = client.post(
+        f"/api/v1.0/rooms/{room.id}/start-recording/",
+        {"mode": "screen_recording"},
+    )
+    assert second.status_code == 201
+    assert Recording.objects.count() == 2
+
+
 def test_start_recording_success(
    mock_worker_service_factory, mock_worker_manager, settings
 ):
@@ -67,7 +67,7 @@ def test_api_rooms_update_members():
            "name": "New name",
            "slug": "should-be-ignored",
            "access_level": RoomAccessLevel.RESTRICTED,
-            "configuration": {"the_key": "the_value"},
+            "configuration": {"can_publish_sources": ["camera", "microphone"]},
        },
        format="json",
    )
@@ -95,7 +95,7 @@ def test_api_rooms_update_administrators():
            "name": "New name",
            "slug": "should-be-ignored",
            "access_level": RoomAccessLevel.PUBLIC,
-            "configuration": {"the_key": "the_value"},
+            "configuration": {"can_publish_sources": ["camera", "microphone"]},
        },
        format="json",
    )
@@ -104,7 +104,98 @@ def test_api_rooms_update_administrators():
    assert room.name == "New name"
    assert room.slug == "new-name"
    assert room.access_level == RoomAccessLevel.PUBLIC
-    assert room.configuration == {"the_key": "the_value"}
+    assert room.configuration == {"can_publish_sources": ["camera", "microphone"]}
+
+
+@pytest.mark.parametrize(
+    "configuration",
+    [
+        {},
+        {"can_publish_sources": ["camera", "microphone"]},
+        {
+            "can_publish_sources": [
+                "camera",
+                "microphone",
+                "screen_share",
+                "screen_share_audio",
+            ]
+        },
+        {"can_publish_sources": []},
+        {"can_publish_sources": None},
+    ],
+)
+def test_api_rooms_update_configuration_valid(configuration):
+    """Administrators should be allowed to set valid configurations."""
+    user = UserFactory()
+    room = RoomFactory(users=[(user, "owner")])
+    client = APIClient()
+    client.force_login(user)
+
+    response = client.patch(
+        f"/api/v1.0/rooms/{room.id!s}/",
+        {"configuration": configuration},
+        format="json",
+    )
+    assert response.status_code == 200
+    room.refresh_from_db()
+    assert room.configuration == configuration
+
+
+def test_api_rooms_update_configuration_extra_keys_rejected():
+    """Extra keys in configuration should be rejected."""
+    user = UserFactory()
+    room = RoomFactory(users=[(user, "owner")])
+    client = APIClient()
+    client.force_login(user)
+
+    response = client.patch(
+        f"/api/v1.0/rooms/{room.id!s}/",
+        {
+            "configuration": {
+                "can_publish_sources": ["camera"],
+                "arbitrary_key": "value",
+            }
+        },
+        format="json",
+    )
+    assert response.status_code == 400
+    room.refresh_from_db()
+    assert room.configuration == {}
+
+
+@pytest.mark.parametrize("invalid_source", ["invalid_source", "CAMERA"])
+def test_api_rooms_update_configuration_invalid_source_value(invalid_source):
+    """Invalid source values should be rejected."""
+    user = UserFactory()
+    room = RoomFactory(users=[(user, "owner")])
+    client = APIClient()
+    client.force_login(user)
+
+    response = client.patch(
+        f"/api/v1.0/rooms/{room.id!s}/",
+        {"configuration": {"can_publish_sources": [invalid_source]}},
+        format="json",
+    )
+    assert response.status_code == 400
+    room.refresh_from_db()
+    assert room.configuration == {}
+
+
+def test_api_rooms_update_configuration_wrong_type():
+    """Configuration values with wrong types should be rejected."""
+    user = UserFactory()
+    room = RoomFactory(users=[(user, "owner")])
+    client = APIClient()
+    client.force_login(user)
+
+    response = client.patch(
+        f"/api/v1.0/rooms/{room.id!s}/",
+        {"configuration": {"can_publish_sources": "camera"}},
+        format="json",
+    )
+    assert response.status_code == 400
+    room.refresh_from_db()
+    assert room.configuration == {}


 def test_api_rooms_update_administrators_of_another():
@@ -576,8 +576,8 @@ msgstr "So speichern Sie diese Aufzeichnung dauerhaft:"

 #: core/templates/mail/html/screen_recording.html:208
 #: core/templates/mail/text/screen_recording.txt:13
-msgid "Click the \"Open\" button below "
-msgstr "Klicken Sie auf den Button „Öffnen“ unten "
+msgid "Click the \"<a href=\"%(link)s\">Open</a>\" link below "
+msgstr "Klicken Sie auf den Link „<a href=\"%(link)s\">Öffnen</a>\" unten "

 #: core/templates/mail/html/screen_recording.html:209
 #: core/templates/mail/text/screen_recording.txt:14
@@ -572,8 +572,8 @@ msgstr "To keep this recording permanently:"

 #: core/templates/mail/html/screen_recording.html:208
 #: core/templates/mail/text/screen_recording.txt:13
-msgid "Click the \"Open\" button below "
-msgstr "Click the \"Open\" button below "
+msgid "Click the \"<a href=\"%(link)s\">Open</a>\" link below "
+msgstr "Click the \"<a href=\"%(link)s\">Open</a>\" link below "

 #: core/templates/mail/html/screen_recording.html:209
 #: core/templates/mail/text/screen_recording.txt:14
@@ -578,8 +578,8 @@ msgstr "Pour conserver cet enregistrement de façon permanente :"

 #: core/templates/mail/html/screen_recording.html:208
 #: core/templates/mail/text/screen_recording.txt:13
-msgid "Click the \"Open\" button below "
-msgstr "Cliquez sur le bouton \"Ouvrir\" ci-dessous "
+msgid "Click the \"<a href=\"%(link)s\">Open</a>\" link below "
+msgstr "Cliquez sur le lien \"<a href=\"%(link)s\">Ouvrir</a>\" ci-dessous "

 #: core/templates/mail/html/screen_recording.html:209
 #: core/templates/mail/text/screen_recording.txt:14
@@ -571,8 +571,8 @@ msgstr "Om deze opname permanent te bewaren:"

 #: core/templates/mail/html/screen_recording.html:208
 #: core/templates/mail/text/screen_recording.txt:13
-msgid "Click the \"Open\" button below "
-msgstr "Klik op de \"Openen\"-knop hieronder "
+msgid "Click the \"<a href=\"%(link)s\">Open</a>\" link below "
+msgstr "Klik op de \"<a href=\"%(link)s\">Openen</a>\"-link hieronder "

 #: core/templates/mail/html/screen_recording.html:209
 #: core/templates/mail/text/screen_recording.txt:14
@@ -700,6 +700,44 @@ class Base(Configuration):
    RECORDING_MAX_DURATION = values.IntegerValue(
        None, environ_name="RECORDING_MAX_DURATION", environ_prefix=None
    )
+
+    # Recording encoding options for LiveKit Egress (video composite egress only).
+    # These settings affect screen recordings handled by VideoCompositeEgressService;
+    # they are silently ignored by AudioCompositeEgressService (audio-only transcript
+    # recordings), whose request never carries advanced EncodingOptions.
+    # When disabled, LiveKit falls back to its built-in H264_720P_30 preset
+    # (1280x720, 30 fps, 3000 kbps H.264 MAIN video, 128 kbps AAC audio).
+    # When enabled, the values below are passed to LiveKit as EncodingOptions
+    # (advanced) and replace the preset. Lowering framerate and bitrate reduces
+    # output file size and CPU load on the egress worker.
+    RECORDING_ENCODING_ENABLED = values.BooleanValue(
+        False, environ_name="RECORDING_ENCODING_ENABLED", environ_prefix=None
+    )
+    RECORDING_ENCODING_WIDTH = values.PositiveIntegerValue(
+        1280, environ_name="RECORDING_ENCODING_WIDTH", environ_prefix=None
+    )
+    RECORDING_ENCODING_HEIGHT = values.PositiveIntegerValue(
+        720, environ_name="RECORDING_ENCODING_HEIGHT", environ_prefix=None
+    )
+    RECORDING_ENCODING_FRAMERATE = values.PositiveIntegerValue(
+        30, environ_name="RECORDING_ENCODING_FRAMERATE", environ_prefix=None
+    )
+    RECORDING_ENCODING_VIDEO_BITRATE_KBPS = values.PositiveIntegerValue(
+        3000,
+        environ_name="RECORDING_ENCODING_VIDEO_BITRATE_KBPS",
+        environ_prefix=None,
+    )
+    RECORDING_ENCODING_AUDIO_BITRATE_KBPS = values.PositiveIntegerValue(
+        128,
+        environ_name="RECORDING_ENCODING_AUDIO_BITRATE_KBPS",
+        environ_prefix=None,
+    )
+    RECORDING_ENCODING_KEY_FRAME_INTERVAL_S = values.FloatValue(
+        4.0,
+        environ_name="RECORDING_ENCODING_KEY_FRAME_INTERVAL_S",
+        environ_prefix=None,
+    )
+
    SUMMARY_SERVICE_ENDPOINT = values.Value(
        None, environ_name="SUMMARY_SERVICE_ENDPOINT", environ_prefix=None
    )
@@ -817,6 +855,11 @@ class Base(Configuration):
        environ_name="METADATA_COLLECTOR_AGENT_NAME",
        environ_prefix=None,
    )
+    METADATA_COLLECTOR_OUTPUT_FOLDER = values.Value(
+        "metadata",
+        environ_name="METADATA_COLLECTOR_OUTPUT_FOLDER",
+        environ_prefix=None,
+    )

    # External Applications
    APPLICATION_ENABLED = values.BooleanValue(
@@ -40,7 +40,7 @@ dependencies = [
    "django-storages[s3]==1.14.6",
    "django-timezone-field>=5.1",
    "django-pydantic-field==0.5.4",
-    "django==5.2.13",
+    "django==5.2.14",
    "djangorestframework==3.16.1",
    "drf_spectacular==0.29.0",
    "dockerflow==2026.3.4",
@@ -61,6 +61,7 @@ dependencies = [
    "mozilla-django-oidc==5.0.2",
    "livekit-api==1.1.0",
    "aiohttp==3.13.4",
+    "urllib3==2.7.0",
 ]

 [project.urls]
@@ -573,16 +573,16 @@ wheels = [

 [[package]]
 name = "django"
-version = "5.2.13"
+version = "5.2.14"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
    { name = "asgiref" },
    { name = "sqlparse" },
    { name = "tzdata", marker = "sys_platform == 'win32'" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/1f/c5/c69e338eb2959f641045802e5ea87ca4bf5ac90c5fd08953ca10742fad51/django-5.2.13.tar.gz", hash = "sha256:a31589db5188d074c63f0945c3888fad104627dfcc236fb2b97f71f89da33bc4", size = 10890368, upload-time = "2026-04-07T14:02:15.072Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/65/95/95f7faa0950867afaa0bef2460c6263afd6a2c78cc9434046ed28160b015/django-5.2.14.tar.gz", hash = "sha256:58a63ba841662e5c686b57ba1fec52ddd68c0b93bd96ac3029d55728f00bf8a2", size = 10895118, upload-time = "2026-05-05T13:57:31.104Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/59/b1/51ab36b2eefcf8cdb9338c7188668a157e29e30306bfc98a379704c9e10d/django-5.2.13-py3-none-any.whl", hash = "sha256:5788fce61da23788a8ce6f02583765ab060d396720924789f97fa42119d37f7a", size = 8310982, upload-time = "2026-04-07T14:02:08.883Z" },
+    { url = "https://files.pythonhosted.org/packages/14/44/f172870cf87aa25afef48fb72adba89ee8b77fcab6f3b23d240b923f1528/django-5.2.14-py3-none-any.whl", hash = "sha256:6f712143bd3064310d1f50fac859c3e9a274bdcfc9595339853be7779297fc76", size = 8311320, upload-time = "2026-05-05T13:57:25.795Z" },
 ]

 [[package]]
@@ -1212,6 +1212,7 @@ dependencies = [
    { name = "redis" },
    { name = "requests" },
    { name = "sentry-sdk" },
+    { name = "urllib3" },
    { name = "whitenoise" },
 ]

@@ -1243,7 +1244,7 @@ requires-dist = [
    { name = "brotli", specifier = "==1.2.0" },
    { name = "celery", extras = ["redis"], specifier = "==5.6.2" },
    { name = "dj-database-url", specifier = "==3.1.2" },
-    { name = "django", specifier = "==5.2.13" },
+    { name = "django", specifier = "==5.2.14" },
    { name = "django-configurations", specifier = "==2.5.1" },
    { name = "django-cors-headers", specifier = "==4.9.0" },
    { name = "django-countries", specifier = "==8.2.0" },
@@ -1273,6 +1274,7 @@ requires-dist = [
    { name = "redis", specifier = "==5.2.1" },
    { name = "requests", specifier = "==2.33.0" },
    { name = "sentry-sdk", specifier = "==2.54.0" },
+    { name = "urllib3", specifier = "==2.7.0" },
    { name = "whitenoise", specifier = "==6.12.0" },
 ]

@@ -2260,11 +2262,11 @@ wheels = [

 [[package]]
 name = "urllib3"
-version = "2.6.3"
+version = "2.7.0"
 source = { registry = "https://pypi.org/simple" }
-sdist = { url = "https://files.pythonhosted.org/packages/c7/24/5f1b3bdffd70275f6661c76461e25f024d5a38a46f04aaca912426a2b1d3/urllib3-2.6.3.tar.gz", hash = "sha256:1b62b6884944a57dbe321509ab94fd4d3b307075e0c2eae991ac71ee15ad38ed", size = 435556, upload-time = "2026-01-07T16:24:43.925Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/53/0c/06f8b233b8fd13b9e5ee11424ef85419ba0d8ba0b3138bf360be2ff56953/urllib3-2.7.0.tar.gz", hash = "sha256:231e0ec3b63ceb14667c67be60f2f2c40a518cb38b03af60abc813da26505f4c", size = 433602, upload-time = "2026-05-07T16:13:18.596Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/39/08/aaaad47bc4e9dc8c725e68f9d04865dbcb2052843ff09c97b08904852d84/urllib3-2.6.3-py3-none-any.whl", hash = "sha256:bf272323e553dfb2e87d9bfd225ca7b0f467b919d7bbd355436d3fd37cb0acd4", size = 131584, upload-time = "2026-01-07T16:24:42.685Z" },
+    { url = "https://files.pythonhosted.org/packages/7f/3e/5db95bcf282c52709639744ca2a8b149baccf648e39c8cc87553df9eae0c/urllib3-2.7.0-py3-none-any.whl", hash = "sha256:9fb4c81ebbb1ce9531cce37674bbc6f1360472bc18ca9a553ede278ef7276897", size = 131087, upload-time = "2026-05-07T16:13:17.151Z" },
 ]

 [[package]]
@@ -0,0 +1,3 @@
+declare module '@fontsource-variable/lexend' {}
+declare module '@fontsource-variable/atkinson-hyperlegible-next' {}
+declare module '@fontsource/opendyslexic' {}
@@ -13,8 +13,11 @@
    "check": "prettier --check ./src"
  },
  "dependencies": {
+    "@fontsource-variable/atkinson-hyperlegible-next": "5.2.6",
+    "@fontsource-variable/lexend": "5.2.11",
    "@fontsource-variable/material-symbols-outlined": "5.2.34",
    "@fontsource/material-icons-outlined": "5.2.6",
+    "@fontsource/opendyslexic": "5.2.5",
    "@livekit/components-react": "2.9.19",
    "@livekit/components-styles": "1.2.0",
    "@livekit/track-processors": "0.7.0",
@@ -44,7 +47,7 @@
    "wouter": "3.9.0"
  },
  "devDependencies": {
-    "@pandacss/dev": "1.8.2",
+    "@pandacss/dev": "1.11.1",
    "@tanstack/eslint-plugin-query": "5.91.4",
    "@tanstack/react-query-devtools": "5.91.3",
    "@types/humanize-duration": "3.27.4",
@@ -59,7 +62,7 @@
    "eslint-plugin-jsx-a11y": "6.10.2",
    "eslint-plugin-react-hooks": "5.2.0",
    "eslint-plugin-react-refresh": "0.4.20",
-    "postcss": "8.5.6",
+    "postcss": "8.5.14",
    "prettier": "3.8.1",
    "typescript": "5.8.3",
    "vite": "7.3.2",
@@ -14,12 +14,14 @@ import './i18n/init'
 import { queryClient } from '@/api/queryClient'
 import { AppInitialization } from '@/components/AppInitialization'
 import { useIsSdkContext } from '@/features/sdk/hooks/useIsSdkContext'
+import { useApplyA11yFonts } from '@/hooks/useApplyA11yFonts'

 function App() {
  const { i18n } = useTranslation()
  useTitle(import.meta.env.VITE_APP_TITLE ?? '')

  const isSDKContext = useIsSdkContext()
+  useApplyA11yFonts()

  return (
    <QueryClientProvider client={queryClient}>
@@ -1,10 +1,20 @@
 // Google Meet uses only letters in a room identifier
 const ROOM_ID_ALLOWED_CHARACTERS = 'abcdefghijklmnopqrstuvwxyz'

-const getRandomChar = () =>
-  ROOM_ID_ALLOWED_CHARACTERS[
-    Math.floor(Math.random() * ROOM_ID_ALLOWED_CHARACTERS.length)
+const getRandomChar = () => {
+  const maxValue =
+    Math.floor(0x100000000 / ROOM_ID_ALLOWED_CHARACTERS.length) *
+    ROOM_ID_ALLOWED_CHARACTERS.length
+  const randomValue = new Uint32Array(1)
+
+  do {
+    crypto.getRandomValues(randomValue)
+  } while (randomValue[0] >= maxValue)
+
+  return ROOM_ID_ALLOWED_CHARACTERS[
+    randomValue[0] % ROOM_ID_ALLOWED_CHARACTERS.length
  ]
+}

 const generateSegment = (length: number): string =>
  Array.from(Array(length), getRandomChar).join('')
@@ -1,9 +1,14 @@
-import { Field, H } from '@/primitives'
+import { Field, H, Text } from '@/primitives'
 import { TabPanel, TabPanelProps } from '@/primitives/Tabs'
 import { css } from '@/styled-system/css'
+import { useMemo } from 'react'
 import { useTranslation } from 'react-i18next'
 import { useSnapshot } from 'valtio'
-import { accessibilityStore } from '@/stores/accessibility'
+import {
+  accessibilityStore,
+  UI_FONT_OPTIONS,
+  type UiFont,
+} from '@/stores/accessibility'
 import { CaptionsSettings } from '@/features/subtitle/component/CaptionsSettings'

 export type AccessibilityTabProps = Pick<TabPanelProps, 'id'>
@@ -12,6 +17,15 @@ export const AccessibilityTab = ({ id }: AccessibilityTabProps) => {
  const { t } = useTranslation('settings')
  const snap = useSnapshot(accessibilityStore)

+  const fontItems = useMemo(
+    () =>
+      UI_FONT_OPTIONS.map((font) => ({
+        value: font,
+        label: t(`accessibility.font.options.${font}`),
+      })),
+    [t]
+  )
+
  return (
    <TabPanel padding={'md'} flex id={id}>
      <H lvl={2}>{t('tabs.accessibility')}</H>
@@ -33,6 +47,21 @@ export const AccessibilityTab = ({ id }: AccessibilityTabProps) => {
            wrapperProps={{ noMargin: true, fullWidth: true }}
          />
        </li>
+        <li>
+          <Field
+            type="select"
+            label={t('accessibility.font.label')}
+            items={fontItems}
+            selectedKey={snap.uiFont}
+            onSelectionChange={(key) => {
+              accessibilityStore.uiFont = key as UiFont
+            }}
+            wrapperProps={{ noMargin: true, fullWidth: true }}
+          />
+          <Text variant="smNote" className={css({ marginTop: '0.25rem' })}>
+            {t('accessibility.font.description')}
+          </Text>
+        </li>
        <CaptionsSettings />
      </ul>
    </TabPanel>
@@ -0,0 +1,38 @@
+import { useEffect } from 'react'
+import { useSnapshot } from 'valtio'
+import { accessibilityStore, UiFont } from '@/stores/accessibility'
+
+const fontImports: Partial<Record<UiFont, () => Promise<unknown>>> = {
+  lexend: () => import('@fontsource-variable/lexend'),
+  'atkinson-hyperlegible': () =>
+    import('@fontsource-variable/atkinson-hyperlegible-next'),
+  opendyslexic: () => import('@fontsource/opendyslexic'),
+}
+
+const loadedFonts = new Set<UiFont>()
+
+export function useApplyA11yFonts() {
+  const { uiFont } = useSnapshot(accessibilityStore)
+
+  useEffect(() => {
+    if (uiFont === 'default') {
+      return
+    }
+
+    const className = `font-${uiFont}`
+    const loader = fontImports[uiFont]
+
+    if (loader && !loadedFonts.has(uiFont)) {
+      loader().then(() => {
+        loadedFonts.add(uiFont)
+        document.documentElement.classList.add(className)
+      })
+    } else {
+      document.documentElement.classList.add(className)
+    }
+
+    return () => {
+      document.documentElement.classList.remove(className)
+    }
+  }, [uiFont])
+}
@@ -1,7 +1,7 @@
 {
  "error": {
    "title": "Aufzeichnung nicht verfügbar",
-    "body": "Die Aufzeichnung ist nicht verfügbar oder wurde möglicherweise gelöscht. Nur die*der Organisator:in der Besprechung hat Zugriff darauf. Wende dich bei Bedarf gerne an die Person."
+    "body": "Die Aufzeichnung ist nicht verfügbar oder wurde möglicherweise gelöscht. Nur der Host der Besprechung hat Zugriff darauf. Wende dich bei Bedarf gerne an die Person."
  },
  "expired": {
    "title": "Aufzeichnung abgelaufen",
@@ -22,7 +22,7 @@
    "button": "Herunterladen",
    "warning": {
      "title": "Linkfreigabe deaktiviert",
-      "body": "Die Freigabe der Aufzeichnung per Link ist noch nicht verfügbar. Nur Organisator:innen können sie herunterladen."
+      "body": "Die Freigabe der Aufzeichnung per Link ist noch nicht verfügbar. Nur Hosts können sie herunterladen."
    }
  }
 }
@@ -3,7 +3,7 @@
    "heading": {
      "normal": "Du hast das Meeting verlassen",
      "duplicateIdentity": "Du bist dem Meeting von einem anderen Gerät aus beigetreten",
-      "participantRemoved": "Du wurdest von der*dem Organisator:in aus dem Meeting entfernt"
+      "participantRemoved": "Du wurdest vom Host aus dem Meeting entfernt"
    },
    "home": "Zur Startseite zurückkehren",
    "back": "Dem Meeting erneut beitreten"
@@ -383,7 +383,7 @@
    "body": "Transkribiere bis zu {{max_duration}} Meetingdauer.",
    "bodyWithoutMaxDuration": "Transkribieren dein Meeting ohne Zeitbegrenzung.",
    "details": {
-      "receiver": "Das Transkript wird an die*den Organisator:in und die Mitorganisierenden gesendet.",
+      "receiver": "Das Transkript wird an den Host und die Co-Hosts gesendet.",
      "destination": "Ein neues Dokument wird erstellt auf",
      "destinationUnknown": "Ein neues Dokument wird erstellt",
      "language": "Meeting-Sprache:",
@@ -438,7 +438,7 @@
    "linkMore": "Dokumentation öffnen",
    "linkAriaLabel": "Dokumentation zur Aufzeichnung öffnen – öffnet in neuem Tab",
    "details": {
-      "receiver": "Die Aufzeichnung wird an die*den Organisator:in und die Mitorganisierenden gesendet.",
+      "receiver": "Die Aufzeichnung wird an den Host und die Co-Hosts gesendet.",
      "destination": "Diese Aufzeichnung wird vorübergehend auf unseren Servern gespeichert",
      "loading": "Aufzeichnung wird gestartet",
      "linkMore": "Mehr erfahren",
@@ -494,7 +494,7 @@
    "button": "OK"
  },
  "admin": {
-    "description": "Diese Einstellungen für Organisierende ermöglichen dir die Kontrolle über dein Meeting. Nur Organisierende haben Zugriff auf diese Optionen.",
+    "description": "Diese Host-Einstellungen ermöglichen dir die Kontrolle über dein Meeting. Nur Hosts haben Zugriff auf diese Optionen.",
    "access": {
      "title": "Raumzugang",
      "description": "Diese Einstellungen gelten auch für zukünftige Meetings in diesem Raum.",
@@ -117,6 +117,16 @@
    "announceReactions": {
      "label": "Reaktionen vorlesen"
    },
+    "font": {
+      "label": "Anzeigeschrift",
+      "description": "Passe die in der Oberfläche verwendete Schrift für ein besseres Leseerlebnis an.",
+      "options": {
+        "default": "Standard (System)",
+        "lexend": "Lexend (verbesserte Lesbarkeit)",
+        "atkinson-hyperlegible": "Atkinson Hyperlegible (Sehbeeinträchtigung)",
+        "opendyslexic": "OpenDyslexic (Dyslexie)"
+      }
+    },
    "captions": {
      "heading": "Untertitel",
      "textSize": {
@@ -1,7 +1,7 @@
 {
  "error": {
    "title": "Recording unavailable",
-    "body": "The recording is unavailable or may have been deleted. Only the meeting organizer can access it. Feel free to contact them if needed."
+    "body": "The recording is unavailable or may have been deleted. Only the meeting host can access it. Feel free to contact them if needed."
  },
  "expired": {
    "title": "Recording expired",
@@ -22,7 +22,7 @@
    "button": "Download",
    "warning": {
      "title": "Link sharing disabled",
-      "body": "Sharing the recording via link is not yet available. Only organizers can download it."
+      "body": "Sharing the recording via link is not yet available. Only hosts can download it."
    }
  }
 }
@@ -3,7 +3,7 @@
    "heading": {
      "normal": "You have left the meeting",
      "duplicateIdentity": "You have joined the meeting from another device",
-      "participantRemoved": "You have been removed from the meeting by an administrator"
+      "participantRemoved": "You have been removed from the meeting by a host"
    },
    "home": "Return to home",
    "back": "Rejoin the meeting"
@@ -328,7 +328,7 @@
      "chat": "Messages in the chat",
      "transcript": "Transcribe",
      "screenRecording": "Record",
-      "admin": "Admin settings",
+      "admin": "Host settings",
      "tools": "More tools",
      "info": "Meeting information"
    },
@@ -338,7 +338,7 @@
      "chat": "messages",
      "transcript": "transcribe",
      "screenRecording": "record",
-      "admin": "admin settings",
+      "admin": "host settings",
      "tools": "more tools",
      "info": "meeting information"
    },
@@ -493,7 +493,7 @@
    "button": "OK"
  },
  "admin": {
-    "description": "These organizer settings allow you to maintain control of your meeting. Only organizers can access these controls.",
+    "description": "These host settings allow you to maintain control of your meeting. Only hosts can access these controls.",
    "access": {
      "title": "Room access",
      "description": "These settings will also apply to future occurrences of this meeting.",
@@ -117,6 +117,16 @@
    "announceReactions": {
      "label": "Announce reactions aloud"
    },
+    "font": {
+      "label": "Display font",
+      "description": "Customize the font used in the interface to improve your reading comfort.",
+      "options": {
+        "default": "Default",
+        "lexend": "Lexend (improved readability)",
+        "atkinson-hyperlegible": "Atkinson Hyperlegible (low vision)",
+        "opendyslexic": "OpenDyslexic (dyslexia)"
+      }
+    },
    "captions": {
      "heading": "Captions",
      "textSize": {
@@ -117,6 +117,16 @@
    "announceReactions": {
      "label": "Vocaliser les réactions"
    },
+    "font": {
+      "label": "Police d'affichage",
+      "description": "Personnalisez la police utilisée dans l'interface pour améliorer votre confort de lecture.",
+      "options": {
+        "default": "Par défaut",
+        "lexend": "Lexend (lisibilité améliorée)",
+        "atkinson-hyperlegible": "Atkinson Hyperlegible (basse vision)",
+        "opendyslexic": "OpenDyslexic (dyslexie)"
+      }
+    },
    "captions": {
      "heading": "Sous-titres",
      "textSize": {
@@ -1,7 +1,7 @@
 {
  "error": {
    "title": "Opname niet beschikbaar",
-    "body": "De opname is niet beschikbaar of is mogelijk verwijderd. Alleen de organisator van de vergadering heeft toegang. Neem gerust contact met hem of haar op als dat nodig is."
+    "body": "De opname is niet beschikbaar of is mogelijk verwijderd. Alleen de host van de vergadering heeft toegang. Neem gerust contact met hem of haar op als dat nodig is."
  },
  "expired": {
    "title": "Opname verlopen",
@@ -22,7 +22,7 @@
    "button": "Downloaden",
    "warning": {
      "title": "Delen via link uitgeschakeld",
-      "body": "Het delen van de opname via een link is nog niet beschikbaar. Alleen organisatoren kunnen deze downloaden."
+      "body": "Het delen van de opname via een link is nog niet beschikbaar. Alleen hosts kunnen deze downloaden."
    }
  }
 }
@@ -437,7 +437,7 @@
    "linkMore": "Documentatie openen",
    "linkAriaLabel": "Documentatie over opname openen - opent in nieuw venster",
    "details": {
-      "receiver": "De opname wordt verzonden naar de organisator en co-organisatoren.",
+      "receiver": "De opname wordt verzonden naar de host en co-hosts.",
      "destination": "Deze opname wordt tijdelijk op onze servers bewaard",
      "loading": "Opname wordt gestart",
      "linkMore": "Meer informatie",
@@ -493,7 +493,7 @@
    "button": "Opnieuw proberen"
  },
  "admin": {
-    "description": "Deze organisatorinstellingen geven u controle over uw vergadering. Alleen organisatoren hebben toegang tot deze bedieningselementen.",
+    "description": "Deze hostinstellingen geven u controle over uw vergadering. Alleen hosts hebben toegang tot deze bedieningselementen.",
    "access": {
      "title": "Toegang tot vergadering",
      "description": "Deze instellingen zijn ook van toepassing op toekomstige sessies van deze vergadering.",
@@ -117,6 +117,16 @@
    "announceReactions": {
      "label": "Reacties hardop aankondigen"
    },
+    "font": {
+      "label": "Weergavelettertype",
+      "description": "Pas het lettertype van de interface aan om uw leescomfort te verbeteren.",
+      "options": {
+        "default": "Standaard",
+        "lexend": "Lexend (verbeterde leesbaarheid)",
+        "atkinson-hyperlegible": "Atkinson Hyperlegible (slechtziend)",
+        "opendyslexic": "OpenDyslexic (dyslexie)"
+      }
+    },
    "captions": {
      "heading": "Ondertitels",
      "textSize": {
@@ -2,6 +2,19 @@ import { proxy, subscribe } from 'valtio'
 import { STORAGE_KEYS } from '@/utils/storageKeys'
 import { deserializeToProxyMap } from '@/utils/valtio'

+export type UiFont =
+  | 'default'
+  | 'lexend'
+  | 'atkinson-hyperlegible'
+  | 'opendyslexic'
+
+export const UI_FONT_OPTIONS: UiFont[] = [
+  'default',
+  'lexend',
+  'atkinson-hyperlegible',
+  'opendyslexic',
+]
+
 export type CaptionTextSize = 'small' | 'medium' | 'large'

 export const CAPTION_TEXT_SIZE_OPTIONS: CaptionTextSize[] = [
@@ -62,6 +75,7 @@ type AccessibilityState = {
  captionTextSize: CaptionTextSize
  captionFontColor: CaptionColor
  captionBackgroundColor: CaptionColor
+  uiFont: UiFont
 }

 const DEFAULT_STATE: AccessibilityState = {
@@ -69,6 +83,7 @@ const DEFAULT_STATE: AccessibilityState = {
  captionTextSize: 'medium',
  captionFontColor: 'default',
  captionBackgroundColor: 'default',
+  uiFont: 'default',
 }

 function getAccessibilityState(): AccessibilityState {
@@ -91,6 +106,9 @@ function getAccessibilityState(): AccessibilityState {
      )
        ? parsed.captionBackgroundColor
        : DEFAULT_STATE.captionBackgroundColor
+      const uiFont = UI_FONT_OPTIONS.includes(parsed.uiFont)
+        ? parsed.uiFont
+        : DEFAULT_STATE.uiFont
      return {
        ...DEFAULT_STATE,
        ...parsed,
@@ -101,6 +119,7 @@ function getAccessibilityState(): AccessibilityState {
        captionTextSize,
        captionFontColor,
        captionBackgroundColor,
+        uiFont,
      }
    }

@@ -6,6 +6,19 @@ body,
  height: 100%;
 }

+html.font-lexend {
+  --fonts-sans: 'Lexend Variable', ui-sans-serif, system-ui, sans-serif;
+}
+
+html.font-atkinson-hyperlegible {
+  --fonts-sans:
+    'Atkinson Hyperlegible Next Variable', ui-sans-serif, system-ui, sans-serif;
+}
+
+html.font-opendyslexic {
+  --fonts-sans: OpenDyslexic, ui-sans-serif, system-ui, sans-serif;
+}
+
 .sr-only {
  position: absolute;
  width: 1px;
@@ -1,52 +1,62 @@
-{{- $envVars := include "meet.common.env" (list . .Values.celeryTranscribe) -}}
-{{- $fullName := include "meet.celeryTranscribe.fullname" . -}}
+{{- $sharedEnvVars := default (dict) .Values.celeryTranscribe.envVars -}}
 {{- $component := "celery-transcribe" -}}
+
+{{- range $idx, $instance := .Values.celeryTranscribe.instances }}
+{{- $instanceValues := merge $instance $.Values.celeryTranscribe }}
+{{- $fullName := printf "%s-%s" (include "meet.celeryTranscribe.fullname" $) $instance.name }}
+{{- $extraInstanceEnvVars := default (dict) $instance.extraEnvVars -}}
+{{- $mergedInstanceEnvVars := merge $extraInstanceEnvVars $sharedEnvVars -}}
+{{- $fakeInstanceObjectForEnvHelper := dict "envVars" $mergedInstanceEnvVars -}}
+{{- $envVars := include "meet.common.env" (list . $fakeInstanceObjectForEnvHelper) -}}
 apiVersion: apps/v1
 kind: Deployment
 metadata:
  name: {{ $fullName }}
  annotations:
-   {{- with .Values.celeryTranscribe.dpAnnotations }}
+   {{- with $instanceValues.dpAnnotations }}
   {{- toYaml . | nindent 4 }}
   {{- end }}
-  namespace: {{ .Release.Namespace | quote }}
+  namespace: {{ $.Release.Namespace | quote }}
  labels:
-    {{- include "meet.common.labels" (list . $component) | nindent 4 }}
+    {{- include "meet.common.labels" (list $ $component) | nindent 4 }}
+    instance: {{ $instance.name }}
 spec:
-  replicas: {{ .Values.celeryTranscribe.replicas }}
+  replicas: {{ $instanceValues.replicas }}
  selector:
    matchLabels:
-      {{- include "meet.common.selectorLabels" (list . $component) | nindent 6 }}
+      {{- include "meet.common.selectorLabels" (list $ $component) | nindent 6 }}
+      instance: {{ $instance.name }}
  template:
    metadata:
      annotations:
-        {{- with .Values.celeryTranscribe.podAnnotations }}
+        {{- with $instanceValues.podAnnotations }}
        {{- toYaml . | nindent 8 }}
        {{- end }}
      labels:
-        {{- include "meet.common.selectorLabels" (list . $component) | nindent 8 }}
+        {{- include "meet.common.selectorLabels" (list $ $component) | nindent 8 }}
+        instance: {{ $instance.name }}
    spec:
      {{- if $.Values.image.credentials }}
      imagePullSecrets:
-        - name: {{ include "meet.secret.dockerconfigjson.name" (dict "fullname" (include "meet.fullname" .) "imageCredentials" $.Values.image.credentials) }}
+        - name: {{ include "meet.secret.dockerconfigjson.name" (dict "fullname" (include "meet.fullname" $) "imageCredentials" $.Values.image.credentials) }}
      {{- end }}
-      shareProcessNamespace: {{ .Values.celeryTranscribe.shareProcessNamespace }}
-      {{- with .Values.celeryTranscribe.podSecurityContext }}
+      shareProcessNamespace: {{ $instanceValues.shareProcessNamespace }}
+      {{- with $instanceValues.podSecurityContext }}
      securityContext:
        {{- toYaml . | nindent 8 }}
      {{- end }}
      containers:
-        {{- with .Values.celeryTranscribe.sidecars }}
+        {{- with $instanceValues.sidecars }}
          {{- toYaml . | nindent 8 }}
        {{- end }}
-        - name: {{ .Chart.Name }}
-          image: "{{ (.Values.celeryTranscribe.image | default dict).repository | default .Values.image.repository }}:{{ (.Values.celeryTranscribe.image | default dict).tag | default .Values.image.tag }}"
-          imagePullPolicy: {{ (.Values.celeryTranscribe.image | default dict).pullPolicy | default .Values.image.pullPolicy }}
-          {{- with .Values.celeryTranscribe.command }}
+        - name: {{ $.Chart.Name }}
+          image: "{{ ($instanceValues.image | default dict).repository | default $.Values.image.repository }}:{{ ($instanceValues.image | default dict).tag | default $.Values.image.tag }}"
+          imagePullPolicy: {{ ($instanceValues.image | default dict).pullPolicy | default $.Values.image.pullPolicy }}
+          {{- with $instanceValues.command }}
          command:
            {{- toYaml . | nindent 12 }}
          {{- end }}
-          {{- with .Values.celeryTranscribe.args }}
+          {{- with $instanceValues.args }}
          args:
            {{- toYaml . | nindent 12 }}
          {{- end }}
@@ -54,65 +64,65 @@ spec:
            {{- if $envVars }}
            {{- $envVars | indent 12 }}
            {{- end }}
-          {{- with .Values.celeryTranscribe.securityContext }}
+          {{- with $instanceValues.securityContext }}
          securityContext:
            {{- toYaml . | nindent 12 }}
          {{- end }}
          ports:
            - name: http
-              containerPort: {{ .Values.celeryTranscribe.service.targetPort }}
+              containerPort: {{ $instanceValues.service.targetPort }}
              protocol: TCP
-          {{- if .Values.celeryTranscribe.probes.liveness }}
+          {{- if $instanceValues.probes.liveness }}
          livenessProbe:
-            {{- include "meet.probes.abstract" (merge .Values.celeryTranscribe.probes.liveness (dict "targetPort" .Values.celeryTranscribe.service.targetPort )) | nindent 12 }}
+            {{- include "meet.probes.abstract" (merge $instanceValues.probes.liveness (dict "targetPort" $instanceValues.service.targetPort)) | nindent 12 }}
          {{- end }}
-          {{- if .Values.celeryTranscribe.probes.readiness }}
+          {{- if $instanceValues.probes.readiness }}
          readinessProbe:
-            {{- include "meet.probes.abstract" (merge .Values.celeryTranscribe.probes.readiness (dict "targetPort" .Values.celeryTranscribe.service.targetPort )) | nindent 12 }}
+            {{- include "meet.probes.abstract" (merge $instanceValues.probes.readiness (dict "targetPort" $instanceValues.service.targetPort)) | nindent 12 }}
          {{- end }}
-          {{- if .Values.celeryTranscribe.probes.startup }}
+          {{- if $instanceValues.probes.startup }}
          startupProbe:
-            {{- include "meet.probes.abstract" (merge .Values.celeryTranscribe.probes.startup (dict "targetPort" .Values.celeryTranscribe.service.targetPort )) | nindent 12 }}
+            {{- include "meet.probes.abstract" (merge $instanceValues.probes.startup (dict "targetPort" $instanceValues.service.targetPort)) | nindent 12 }}
          {{- end }}
-          {{- with .Values.celeryTranscribe.resources }}
+          {{- with $instanceValues.resources }}
          resources:
            {{- toYaml . | nindent 12 }}
          {{- end }}
          volumeMounts:
-            {{- range $index, $value := .Values.mountFiles }}
+            {{- range $index, $value := $.Values.mountFiles }}
            - name: "files-{{ $index }}"
              mountPath: {{ $value.path }}
              subPath: content
            {{- end }}
-            {{- range $name, $volume := .Values.celeryTranscribe.persistence }}
+            {{- range $name, $volume := $instanceValues.persistence }}
            - name: "{{ $name }}"
              mountPath: "{{ $volume.mountPath }}"
            {{- end }}
-            {{- range .Values.celeryTranscribe.extraVolumeMounts }}
+            {{- range $instanceValues.extraVolumeMounts }}
            - name: {{ .name }}
              mountPath: {{ .mountPath }}
              subPath: {{ .subPath | default "" }}
              readOnly: {{ .readOnly }}
            {{- end }}
-      {{- with .Values.celeryTranscribe.nodeSelector }}
+      {{- with $instanceValues.nodeSelector }}
      nodeSelector:
        {{- toYaml . | nindent 8 }}
      {{- end }}
-      {{- with .Values.celeryTranscribe.affinity }}
+      {{- with $instanceValues.affinity }}
      affinity:
        {{- toYaml . | nindent 8 }}
      {{- end }}
-      {{- with .Values.celeryTranscribe.tolerations }}
+      {{- with $instanceValues.tolerations }}
      tolerations:
        {{- toYaml . | nindent 8 }}
      {{- end }}
      volumes:
-        {{- range $index, $value := .Values.mountFiles }}
+        {{- range $index, $value := $.Values.mountFiles }}
        - name: "files-{{ $index }}"
          configMap:
            name: "{{ include "meet.fullname" $ }}-files-{{ $index }}"
        {{- end }}
-        {{- range $name, $volume := .Values.celeryTranscribe.persistence }}
+        {{- range $name, $volume := $instanceValues.persistence }}
        - name: "{{ $name }}"
          {{- if eq $volume.type "emptyDir" }}
          emptyDir: {}
@@ -121,7 +131,7 @@ spec:
            claimName: "{{ $fullName }}-{{ $name }}"
          {{- end }}
        {{- end }}
-        {{- range .Values.celeryTranscribe.extraVolumes }}
+        {{- range $instanceValues.extraVolumes }}
        - name: {{ .name }}
          {{- if .existingClaim }}
          persistentVolumeClaim:
@@ -143,15 +153,18 @@ spec:
          {{- end }}
        {{- end }}
 ---
-{{ if .Values.celeryTranscribe.pdb.enabled }}
+{{ if $instanceValues.pdb.enabled }}
 apiVersion: policy/v1
 kind: PodDisruptionBudget
 metadata:
  name: {{ $fullName }}
-  namespace: {{ .Release.Namespace | quote }}
+  namespace: {{ $.Release.Namespace | quote }}
 spec:
  maxUnavailable: 1
  selector:
    matchLabels:
-      {{- include "meet.common.selectorLabels" (list . $component) | nindent 6 }}
+      {{- include "meet.common.selectorLabels" (list $ $component) | nindent 6 }}
+      instance: {{ $instance.name }}
+---
 {{ end }}
+{{- end }}
@@ -848,6 +848,21 @@ celeryTranscribe:
  pdb:
    enabled: false

+  ## @param celeryTranscribe.instances List of celeryTranscribe instances to deploy. Each entry creates a dedicated Deployment. Useful when wanting to use multiple instances of WhisperX (configure each endpoint in the extraEnv value specific to that instance).
+  ## @extra celeryTranscribe.instances[].name Unique name suffix for the instance (used in the Deployment name and pod labels)
+  ## @extra celeryTranscribe.instances[].replicas Override the number of replicas for this specific instance
+  ## @extra celeryTranscribe.instances[].extraEnvVars Additional environment variables for this specific instance (same structure as envVars)
+  ## @extra celeryTranscribe.instances[].command Override the container command for this specific instance
+  ## @extra celeryTranscribe.instances[].args Override the container args for this specific instance
+  ## @extra celeryTranscribe.instances[].resources Override resource requirements for this specific instance
+  ## @extra celeryTranscribe.instances[].nodeSelector Override node selector for this specific instance
+  ## @extra celeryTranscribe.instances[].affinity Override affinity for this specific instance
+  ## @extra celeryTranscribe.instances[].tolerations Override tolerations for this specific instance
+  ## @extra celeryTranscribe.instances[].pdb.enabled Enable pdb for this specific instance
+  instances:
+    - name: default
+      extraEnvVars: {}
+
 ## @section celerySummarize

 celerySummarize:
@@ -37,7 +37,7 @@
          <mj-text>
            <p>{% trans "To keep this recording permanently:" %}</p>
            <ol>
-              <li>{% blocktrans %}Click the "Open" button below {% endblocktrans %}</li>
+              <li>{% blocktrans %}Click the "<a href="{{link}}">Open</a>" link below {% endblocktrans %}</li>
              <li>{% blocktrans %}Use the "Download" button in the interface {% endblocktrans %}</li>
              <li>{% blocktrans %}Save the file to your preferred location{% endblocktrans %}</li>
            </ol>
@@ -5,7 +5,7 @@
  "type": "module",
  "dependencies": {
    "@html-to/text-cli": "0.5.4",
-    "mjml": "4.18.0"
+    "mjml": "5.0.0"
  },
  "private": true,
  "scripts": {
@@ -20,7 +20,7 @@ dependencies = [
 [project.optional-dependencies]
 dev = [
    "ruff==0.15.6",
-    "pytest==9.0.2",
+    "pytest==9.0.3",
    "responses>=0.25.8",
 ]

@@ -19,16 +19,18 @@ class TranscribeSummarizeTaskCreation(BaseModel):
    """Transcription and summarization parameters."""

    owner_id: str
-    filename: str
+    recording_filename: str
+    metadata_filename: Optional[str] = None
    email: str
    sub: str
    version: Optional[int] = 2
    room: Optional[str]
-    recording_date: Optional[str]
-    recording_time: Optional[str]
+    owner_timezone: Optional[str]
    language: Optional[str]
    download_link: Optional[str]
    context_language: Optional[str] = None
+    recording_start_at: Optional[str] = None
+    recording_end_at: Optional[str] = None

    @field_validator("language")
    @classmethod
@@ -51,16 +53,18 @@ async def create_transcribe_summarize_task(request: TranscribeSummarizeTaskCreat
    task = process_audio_transcribe_summarize_v2.apply_async(
        args=[
            request.owner_id,
-            request.filename,
+            request.recording_filename,
+            request.metadata_filename,
            request.email,
            request.sub,
            time.time(),
            request.room,
-            request.recording_date,
-            request.recording_time,
+            request.owner_timezone,
            request.language,
            request.download_link,
            request.context_language,
+            request.recording_start_at,
+            request.recording_end_at,
        ],
        queue=settings.transcribe_queue,
    )
@@ -112,7 +112,9 @@ class MetadataManager:
        if self._is_disabled or self.has_task_id(task_id):
            return

-        _, filename, email, _, received_at, *_ = task_args
+        # Positional args mirror process_audio_transcribe_summarize_v2 signature:
+        # owner_id, recording_filename, metadata_filename, email, sub, received_at, ...
+        _, filename, _, email, _, received_at, *_ = task_args

        start_time = time.time()
        initial_metadata = {
@@ -4,7 +4,7 @@

 import json
 import time
-from typing import Optional
+from datetime import datetime

 import openai
 import sentry_sdk
@@ -40,6 +40,7 @@ from summary.core.shared_models import (
    webhook_payload_adapter,
 )
 from summary.core.transcript_formatter import TranscriptFormatter
+from summary.core.user_assign import resolve_speaker_identities
 from summary.core.webhook_service import (
    call_webhook_v2,
    submit_content,
@@ -78,7 +79,7 @@ file_service = FileService()
 def transcribe_audio(
    *,
    task_id: str,
-    filename: str | None = None,
+    recording_filename: str | None = None,
    language: str,
    cloud_storage_url=None,
    raises: bool = False,
@@ -90,7 +91,7 @@ def transcribe_audio(

    Returns the transcription object, or None if the file could not be retrieved.
    """
-    if bool(filename) == bool(cloud_storage_url):
+    if bool(recording_filename) == bool(cloud_storage_url):
        raise ValueError(
            "Either filename or cloud_storage_url must be provided, but not both."
        )
@@ -105,11 +106,12 @@ def transcribe_audio(
    # Transcription
    try:
        with file_service.prepare_audio_file(
-            remote_object_key=filename,
+            remote_object_key=recording_filename,
            cloud_storage_url=cloud_storage_url,
        ) as (audio_file, metadata):
            metadata_manager.track(task_id, {"audio_length": metadata["duration"]})

+            # Compute language parameter
            if language is None:
                language = settings.whisperx_default_language
                logger.info(
@@ -122,18 +124,21 @@ def transcribe_audio(
                    language,
                )

+            # Call remote service for transcription
            transcription_start_time = time.time()
-
            transcription = whisperx_client.audio.transcriptions.create(
                model=settings.whisperx_asr_model, file=audio_file, language=language
            )

-            transcription_time = round(time.time() - transcription_start_time, 2)
+            # Logging
+            transcription_duration = round(time.time() - transcription_start_time, 2)
            metadata_manager.track(
                task_id,
-                {"transcription_time": transcription_time},
+                {"transcription_time": transcription_duration},
+            )
+            logger.info(
+                "Transcription received in %.2f seconds.", transcription_duration
            )
-            logger.info("Transcription received in %.2f seconds.", transcription_time)
            logger.debug("Transcription: \n %s", transcription)

    except FileServiceException as e:
@@ -148,7 +153,7 @@ def transcribe_audio(
                "Unexpected error while preparing file | filename: %s "
                "| cloud_storage_url: %s"
            ),
-            filename,
+            recording_filename,
            redacted_cloud_storage_url,
        )
        return None
@@ -157,13 +162,72 @@ def transcribe_audio(
    return transcription


+def resolve_speaker_identities_and_apply_to(
+    transcription, recording_start_at, recording_end_at, metadata_filename, task_id
+):
+    """Assign users to detected speakers and rewrite the transcriptions.
+
+    Args:
+        transcription: output of meet-whisperx after transcription and diarization
+        recording_start_at: sourced from LiveKit FileInfo via the egress_ended webhook
+        recording_end_at: sourced from LiveKit FileInfo via the egress_ended webhook
+        metadata_filename: name of metadata file containing VAD information in S3
+        task_id: current task id, for logging purposes
+    """
+    recording_start_dt = (
+        datetime.fromisoformat(recording_start_at) if recording_start_at else None
+    )
+    recording_end_dt = (
+        datetime.fromisoformat(recording_end_at) if recording_end_at else None
+    )
+
+    logger.debug(
+        "recording_start_dt: %s ; recording_end_dt: %s",
+        recording_start_dt,
+        recording_end_dt,
+    )
+    if (recording_start_dt is None) or (recording_end_dt is None):
+        logger.debug("Skipping resolve_speaker_identities")
+        return transcription
+
+    logger.debug("Running resolve_speaker_identities")
+    try:
+        metadata = file_service.read_json(metadata_filename)
+        speaker_mapping = resolve_speaker_identities(
+            metadata,
+            transcription,
+            recording_start_dt,
+            recording_end_dt,
+        )
+        new_transcription = speaker_mapping.apply_to(transcription.model_dump())
+        return new_transcription
+
+    except FileServiceException as exc:
+        logger.error(
+            "Error reading metadata for task %s; skipping speaker assignment."
+            " Error: %s",
+            task_id,
+            exc,
+        )
+        return transcription
+
+    except Exception as exc:
+        logger.exception(
+            "resolve_speaker_identities failed for task %s; skipping"
+            " speaker assignment. Error: %s",
+            task_id,
+            exc,
+        )
+        return transcription
+
+
 def format_transcript(
    transcription,
    context_language: str | None,
    language: str,
    room: str | None,
-    recording_date: str | None,
-    recording_time: str | None,
+    recording_datetime: str | None,
+    owner_timezone: str | None,
    download_link: str | None,
 ) -> tuple[str, str]:
    """Format a transcription into readable content with a title.
@@ -179,8 +243,8 @@ def format_transcript(
    return formatter.format(
        transcription,
        room=room,
-        recording_date=recording_date,
-        recording_time=recording_time,
+        recording_datetime=recording_datetime,
+        owner_timezone=owner_timezone,
        download_link=download_link,
    )

@@ -188,7 +252,7 @@ def format_transcript(
 def format_actions(llm_output: dict) -> str:
    """Format the actions from the LLM output into a markdown list.

-    fomat:
+    format:
    - [ ] Action title Assignée à : assignee1, assignee2, Échéance : due_date
    """
    lines = []
@@ -212,16 +276,18 @@ def format_actions(llm_output: dict) -> str:
 def process_audio_transcribe_summarize_v2(
    self,
    owner_id: str,
-    filename: str,
+    recording_filename: str,
+    metadata_filename: str | None,
    email: str,
    sub: str,
    received_at: float,
-    room: Optional[str],
-    recording_date: Optional[str],
-    recording_time: Optional[str],
-    language: Optional[str],
-    download_link: Optional[str],
-    context_language: Optional[str] = None,
+    room: str | None,
+    owner_timezone: str | None,
+    language: str | None,
+    download_link: str | None,
+    context_language: str | None = None,
+    recording_start_at: str | None = None,
+    recording_end_at: str | None = None,
 ):
    """Process an audio file by transcribing it and generating a summary.

@@ -234,16 +300,20 @@ def process_audio_transcribe_summarize_v2(
    Args:
        self: Celery task instance (passed on with bind=True)
        owner_id: Unique identifier of the recording owner.
-        filename: Name of the audio file in MinIO storage.
+        recording_filename: Name of the audio file in MinIO storage.
+        metadata_filename: Name of the audio file in MinIO storage.
        email: Email address of the recording owner.
        sub: OIDC subject identifier of the recording owner.
        received_at: Unix timestamp when the recording was received.
        room: room name where the recording took place.
-        recording_date: Date of the recording (localized display string).
-        recording_time: Time of the recording (localized display string).
+        owner_timezone: IANA timezone of the recording owner (e.g. "Europe/Paris").
        language: ISO 639-1 language code for transcription.
        download_link: URL to download the original recording.
        context_language: ISO 639-1 language code of the meeting summary context text.
+        recording_start_at: ISO 8601 timestamp of when file recording actually started
+            (from LiveKit FileInfo.started_at via the egress_ended webhook).
+        recording_end_at: ISO 8601 timestamp of when file recording ended
+            (from LiveKit FileInfo.ended_at via the egress_ended webhook).
    """
    logger.info(
        "Notification received | Owner: %s | Room: %s",
@@ -253,19 +323,33 @@ def process_audio_transcribe_summarize_v2(

    task_id = self.request.id

+    # Transcribe the audio
    transcription = transcribe_audio(
-        task_id=task_id, filename=filename, language=language
+        task_id=task_id, recording_filename=recording_filename, language=language
    )
    if transcription is None:
        return

+    # Assign speakers and rewrite transcription/diarization output
+    if settings.is_resolve_speaker_identities_enabled and (
+        metadata_filename is not None
+    ):
+        transcription = resolve_speaker_identities_and_apply_to(
+            transcription,
+            recording_start_at,
+            recording_end_at,
+            metadata_filename,
+            task_id,
+        )
+
+    # Format output
    content, title = format_transcript(
        transcription,
        context_language,
        language,
        room,
-        recording_date,
-        recording_time,
+        recording_start_at,
+        owner_timezone,
        download_link,
    )

@@ -103,6 +103,12 @@ class Settings(BaseSettings):
    # Transcription processing
    hallucination_patterns: List[str] = ["Vap'n'Roll Thierry"]

+    # Speaker to user assignment
+    is_resolve_speaker_identities_enabled: bool = True
+    resolve_speaker_identities_default_overlap_threshold: float = 0.5
+    resolve_speaker_identities_enable_split_on_words: bool = True
+    resolve_speaker_identities_max_word_duration: float = 1  # seconds
+
    # Webhook-related settings
    webhook_max_retries: int = 2
    webhook_status_forcelist: List[int] = [502, 503, 504]
@@ -23,6 +23,97 @@ settings = get_settings()
 logger = logging.getLogger(__name__)


+def _get_duration_from_packets(local_path: Path) -> float:
+    """Estimate duration from audio packet timestamps."""
+    # Run ffprobe to inspect the first audio stream in the file.
+    # ffprobe is part of FFmpeg and can output media metadata as JSON.
+    #
+    # ruff: noqa: S607 Hard to know the ffprobe path, it depends on the deployment
+    result = subprocess.run(
+        [
+            "ffprobe",
+            # Suppress normal ffprobe logging output.
+            "-v",
+            "quiet",
+            # Ask ffprobe to return JSON.
+            "-print_format",
+            "json",
+            # Select only the first audio stream.
+            "-select_streams",
+            "a:0",
+            # Include packet-level information in the output.
+            "-show_packets",
+            # Only include each packet's start timestamp and duration.
+            "-show_entries",
+            "packet=pts_time,duration_time",
+            # Read only the last ~10 packets 99999999 is to go to the end of the file
+            "-read_intervals",
+            "99999999%+#10",
+            # Skip non-reference frames for speed
+            "-skip_frame",
+            "noref",
+            local_path,
+        ],
+        check=False,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        text=True,  # Decode stdout/stderr as strings instead of bytes.
+    )
+
+    data = json.loads(result.stdout)
+    # Build a list containing the end time of each audio packet.
+    #
+    # For each packet:
+    #   end time = packet start time + packet duration
+    #
+    # pts_time is the packet presentation timestamp, meaning when that packet
+    # starts during playback.
+    #
+    # duration_time may be missing, so it defaults to 0.
+    packet_ends = [
+        float(packet["pts_time"]) + float(packet.get("duration_time", 0))
+        for packet in data.get("packets", [])
+        if "pts_time" in packet
+    ]
+
+    # If no usable packets were found, the duration cannot be estimated.
+    if not packet_ends:
+        raise ValueError("Unable to determine recording duration.")
+
+    # The recording duration is estimated as the latest packet end time.
+    return max(packet_ends)
+
+
+def get_media_duration(local_path: Path):
+    """Get media (audio or video) file duration in seconds."""
+    # ruff: noqa: S607 Hard to know the ffprobe path, it depends on the deployment
+    result = subprocess.run(
+        [
+            "ffprobe",
+            # Suppress normal ffprobe logging output.
+            "-v",
+            "quiet",
+            # Ask ffprobe to return JSON.
+            "-print_format",
+            "json",
+            "-show_format",
+            local_path,
+        ],
+        check=False,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        text=True,
+    )
+
+    data = json.loads(result.stdout)
+    duration_value = data.get("format", {}).get("duration")
+
+    if duration_value not in (None, "N/A"):
+        return float(duration_value)
+
+    return _get_duration_from_packets(local_path)
+
+
 class FileServiceException(Exception):
    """Base exception for file service operations."""

@@ -155,25 +246,7 @@ class FileService:

    def _validate_duration(self, local_path: Path) -> float:
        """Validate audio file duration against configured maximum."""
-        # ruff: noqa: S607 Hard to know the ffprobe path, it depends on the deployment
-        result = subprocess.run(
-            [
-                "ffprobe",
-                "-v",
-                "quiet",
-                "-print_format",
-                "json",
-                "-show_format",
-                local_path,
-            ],
-            check=False,
-            stdout=subprocess.PIPE,
-            stderr=subprocess.PIPE,
-            text=True,
-        )
-
-        data = json.loads(result.stdout)
-        duration = float(data["format"]["duration"])
+        duration = get_media_duration(local_path)

        logger.info(
            "Recording file duration: %.2f seconds",
@@ -229,6 +302,28 @@ class FileService:
                os.remove(output_path)
            raise RuntimeError("Failed to extract audio.") from e

+    def read_json(self, object_name: str) -> dict:
+        """Read and parse a JSON file from MinIO storage."""
+        logger.info("Reading JSON: %s", object_name)
+
+        if not object_name:
+            raise ValueError("Invalid object_name")
+
+        response = None
+        try:
+            response = self._minio_client.get_object(self._bucket_name, object_name)
+            return json.loads(response.read())
+        except (MinioException, S3Error) as e:
+            raise FileServiceException(
+                "Unexpected error while reading JSON object."
+            ) from e
+        except (json.JSONDecodeError, UnicodeDecodeError) as e:
+            raise FileServiceException("Invalid JSON content.") from e
+        finally:
+            if response:
+                response.close()
+                response.release_conn()
+
    @contextmanager
    def prepare_audio_file(
        self,
@@ -1,7 +1,9 @@
 """Transcript formatting into readable conversation format with speaker labels."""

 import logging
-from typing import Optional, Tuple
+from datetime import datetime
+from typing import Tuple
+from zoneinfo import ZoneInfo

 from summary.core.config import get_settings
 from summary.core.locales import LocaleStrings
@@ -39,10 +41,10 @@ class TranscriptFormatter:
    def format(
        self,
        transcription,
-        room: Optional[str] = None,
-        recording_date: Optional[str] = None,
-        recording_time: Optional[str] = None,
-        download_link: Optional[str] = None,
+        room: str | None = None,
+        recording_datetime: str | None = None,
+        owner_timezone: str | None = None,
+        download_link: str | None = None,
    ) -> Tuple[str, str]:
        """Format transcription into the final document and its title."""
        segments = self._get_segments(transcription)
@@ -54,7 +56,7 @@ class TranscriptFormatter:
            content = self._remove_hallucinations(content)
            content = self._add_header(content, download_link)

-        title = self._generate_title(room, recording_date, recording_time)
+        title = self._generate_title(room, recording_datetime, owner_timezone)

        return content, title

@@ -83,7 +85,7 @@ class TranscriptFormatter:

        return formatted_output

-    def _add_header(self, content, download_link: Optional[str]) -> str:
+    def _add_header(self, content, download_link: str | None) -> str:
        """Add download link header to the document content."""
        if not download_link:
            return content
@@ -97,16 +99,20 @@ class TranscriptFormatter:

    def _generate_title(
        self,
-        room: Optional[str] = None,
-        recording_date: Optional[str] = None,
-        recording_time: Optional[str] = None,
+        room: str | None = None,
+        recording_datetime: str | None = None,
+        owner_timezone: str | None = None,
    ) -> str:
        """Generate title from context or return default."""
-        if not room or not recording_date or not recording_time:
+        if not room or not recording_datetime:
            return self._locale.document_default_title

+        dt = datetime.fromisoformat(recording_datetime)
+        if owner_timezone:
+            dt = dt.astimezone(ZoneInfo(owner_timezone))
+
        return self._locale.document_title_template.format(
            room=room,
-            room_recording_date=recording_date,
-            room_recording_time=recording_time,
+            room_recording_date=dt.strftime("%Y-%m-%d"),
+            room_recording_time=dt.strftime("%H:%M"),
        )
@@ -0,0 +1,400 @@
+"""Assign WhisperX diarization speakers to participant identities.
+
+Uses per-stream VAD events to match generic SPEAKER_XX labels provided
+by diarization to real user id's by computing time overlap between
+diarization segments and VAD intervals.
+
+Multiple speakers can map to the same participant (e.g. two people sharing
+one microphone). A participant with no matching speaker gets no assignment.
+"""
+
+import logging
+from collections import defaultdict
+from dataclasses import dataclass, field
+from datetime import datetime
+from typing import Any
+
+from summary.core.config import get_settings
+
+settings = get_settings()
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class Interval:
+    """A time interval in seconds relative to recording start."""
+
+    start: float
+    end: float
+
+
+@dataclass
+class SpeakerAssignment:
+    """Maps a diarization speaker label to a participant."""
+
+    speaker_label: str
+    participant_id: str
+    participant_name: str
+    score: float
+
+
+@dataclass
+class AssignmentResult:
+    """Result of speaker-to-participant assignment."""
+
+    assignments: list[SpeakerAssignment] = field(default_factory=list)
+    unassigned_speakers: list[str] = field(default_factory=list)
+
+    def apply_to(self, diarization: dict[str, Any]) -> dict[str, Any]:
+        """Return a copy of diarization with speaker labels replaced by names.
+
+        Replaces `"speaker"` fields in segments and word_segments with the
+        assigned participant name.  Unassigned speakers are left as-is.
+
+        Args:
+            diarization: WhisperX dict with `segments` and optionally
+                `word_segments`.
+
+        Returns:
+            New dict with speaker labels replaced.
+        """
+        speaker_to_name = {
+            a.speaker_label: a.participant_name for a in self.assignments
+        }
+
+        name_to_speaker_count = defaultdict(int)
+        for name in speaker_to_name.values():
+            name_to_speaker_count[name] += 1
+
+        def _replace_speaker(item: dict[str, Any]) -> dict[str, Any]:
+            if "speaker" in item and item["speaker"] in speaker_to_name:
+                name = speaker_to_name[item["speaker"]]
+                suffix = (
+                    f" ({item['speaker']})" if name_to_speaker_count[name] > 1 else ""
+                )  # Add suffix only if there are multiple detected speakers per user
+                return {**item, "speaker": f"{name}{suffix}"}
+            return {**item}
+
+        def _process_segment(
+            item: dict[str, Any], include_words: bool = False
+        ) -> dict[str, Any]:
+            new_item = _replace_speaker(item)
+            if include_words and "words" in item:
+                new_item["words"] = [_replace_speaker(w) for w in item["words"]]
+            return new_item
+
+        result: dict[str, Any] = {}
+        for key, value in diarization.items():
+            if key not in ("segments", "word_segments"):
+                result[key] = value
+                continue
+            result[key] = [
+                _process_segment(item, include_words=(key == "segments"))
+                for item in value
+            ]
+        return result
+
+
+def _merge_intervals(intervals: list[Interval]) -> list[Interval]:
+    """Return a list of non-overlapping intervals sorted by start time."""
+    if not intervals:
+        return []
+    sorted_intervals = sorted(intervals, key=lambda interval: interval.start)
+    merged: list[Interval] = [
+        Interval(sorted_intervals[0].start, sorted_intervals[0].end)
+    ]
+    for interval in sorted_intervals[1:]:
+        if interval.start <= merged[-1].end:
+            merged[-1].end = max(merged[-1].end, interval.end)
+        else:
+            merged.append(Interval(interval.start, interval.end))
+    return merged
+
+
+def _total_duration(intervals: list[Interval]) -> float:
+    """Return the sum of all interval durations."""
+    return sum(interval.end - interval.start for interval in intervals)
+
+
+def _overlap_duration(
+    a_intervals: list[Interval],
+    b_intervals: list[Interval],
+) -> float:
+    """Compute total overlap between two merged interval lists, sorted by start time."""
+    overlap = 0.0
+    i = j = 0
+    while i < len(a_intervals) and j < len(b_intervals):
+        a = a_intervals[i]
+        b = b_intervals[j]
+        lo = max(a.start, b.start)
+        hi = min(a.end, b.end)
+        if lo < hi:
+            overlap += hi - lo
+        if a.end <= b.end:
+            i += 1
+        else:
+            j += 1
+    return overlap
+
+
+def _format_timelines_debug(
+    participant_timelines: dict[str, list[Interval]],
+    participant_names: dict[str, str],
+    speaker_timelines: dict[str, list[Interval]],
+) -> str:
+    """Render participant and speaker timelines side-by-side for debugging.
+
+    Each row is the slice between two consecutive interval boundaries
+    (drawn from both sides). A filled cell marks an active participant
+    (left block) or speaker (right block) during that slice, so vertical
+    alignment makes overlap visually obvious.
+    """
+    participant_ids = sorted(participant_timelines.keys())
+    speaker_labels = sorted(speaker_timelines.keys())
+
+    if not participant_ids and not speaker_labels:
+        return "(no timelines)"
+
+    boundaries: set[float] = set()
+    for intervals in (*participant_timelines.values(), *speaker_timelines.values()):
+        for iv in intervals:
+            boundaries.add(iv.start)
+            boundaries.add(iv.end)
+    sorted_boundaries = sorted(boundaries)
+    if len(sorted_boundaries) < 2:
+        return "(no intervals)"
+
+    p_headers = [participant_names.get(pid, pid) for pid in participant_ids]
+    s_headers = list(speaker_labels)
+    p_widths = [max(len(h), 3) for h in p_headers]
+    s_widths = [max(len(h), 3) for h in s_headers]
+
+    def _active(intervals: list[Interval], lo: float, hi: float) -> bool:
+        mid = (lo + hi) / 2
+        return any(iv.start <= mid < iv.end for iv in intervals)
+
+    def _cells(
+        intervals_list: list[list[Interval]],
+        widths: list[int],
+        lo: float,
+        hi: float,
+    ) -> str:
+        return " ".join(
+            ("█" * w if _active(iv, lo, hi) else "·" * w)
+            for iv, w in zip(intervals_list, widths, strict=True)
+        )
+
+    p_iv_list = [participant_timelines[pid] for pid in participant_ids]
+    s_iv_list = [speaker_timelines[sl] for sl in speaker_labels]
+
+    time_col = "[   start →      end]"
+    p_hdr = (
+        " ".join(h.center(w) for h, w in zip(p_headers, p_widths, strict=True))
+        or "(none)"
+    )
+    s_hdr = (
+        " ".join(h.center(w) for h, w in zip(s_headers, s_widths, strict=True))
+        or "(none)"
+    )
+    sep = "  ||  "
+    lines = [
+        f"{time_col}  {p_hdr}{sep}{s_hdr}",
+        "-" * (len(time_col) + 2 + len(p_hdr) + len(sep) + len(s_hdr)),
+    ]
+    for lo, hi in zip(sorted_boundaries, sorted_boundaries[1:], strict=False):
+        time_str = f"[{lo:8.2f} → {hi:8.2f}]"
+        p_row = _cells(p_iv_list, p_widths, lo, hi) or " " * len(p_hdr)
+        s_row = _cells(s_iv_list, s_widths, lo, hi) or " " * len(s_hdr)
+        lines.append(f"{time_str}  {p_row}{sep}{s_row}")
+    return "\n".join(lines)
+
+
+def _build_participant_timelines(
+    metadata: dict[str, Any],
+    recording_start_datetime: datetime,
+    recording_end_datetime: datetime | None = None,
+) -> tuple[dict[str, list[Interval]], dict[str, str]]:
+    """Build VAD interval timelines for each participant.
+
+    Args:
+        metadata: Dict with `events` and `participants` keys.
+        recording_start_datetime: UTC datetime used as t=0 reference.
+        recording_end_datetime: UTC datetime of recording end. When provided,
+            any open speech_start without a matching speech_end is closed at
+            this time (the participant is assumed to be speaking until the end).
+
+    Returns:
+        participant_id → merged VAD intervals
+            (seconds relative to recording_start_datetime).
+        participant_id → display name.
+        Intervals are in seconds relative to recording_start_datetime.
+        Events before recording start are clamped to 0.
+    """
+    events = metadata.get("events", [])
+    participants_info = {
+        p["participantId"]: p.get("name", p["participantId"])
+        for p in metadata.get("participants", [])
+    }
+
+    ref_epoch = recording_start_datetime.timestamp()
+
+    open_starts: dict[str, float] = {}
+    intervals: dict[str, list[Interval]] = {}
+
+    for event in events:
+        pid = event["participant_id"]
+        ts = datetime.fromisoformat(event["timestamp"]).timestamp() - ref_epoch
+        etype = event["type"]
+
+        if etype == "speech_start":
+            open_starts[pid] = max(ts, 0.0)
+        elif etype == "speech_end":
+            start = open_starts.pop(pid, None)
+            if start is not None:
+                end = max(ts, 0.0)
+                if end > start:
+                    intervals.setdefault(pid, []).append(Interval(start, end))
+
+    # Close any speech_start that was never matched by a speech_end.
+    # Assume the participant kept speaking until the recording ended.
+    if recording_end_datetime is not None and open_starts:
+        recording_end = recording_end_datetime.timestamp() - ref_epoch
+        for pid, start in open_starts.items():
+            end = max(recording_end, 0.0)
+            if end > start:
+                intervals.setdefault(pid, []).append(Interval(start, end))
+
+    for pid, pid_intervals in intervals.items():
+        intervals[pid] = _merge_intervals(pid_intervals)
+
+    return intervals, participants_info
+
+
+def _build_speaker_timelines(transcription: Any) -> dict[str, list[Interval]]:
+    """Build interval timelines from WhisperX transcription segments."""
+    intervals: dict[str, list[Interval]] = {}
+    segments = transcription.segments if hasattr(transcription, "segments") else []
+    max_word_duration = settings.resolve_speaker_identities_max_word_duration
+
+    for segment in segments:
+        speaker = segment.get("speaker")
+        if speaker is None:
+            continue
+
+        words = [
+            w
+            for w in segment.get("words", [])
+            if w.get("start") is not None and w.get("end") is not None
+        ]
+        if not words:
+            intervals.setdefault(speaker, []).append(
+                Interval(segment["start"], segment["end"])
+            )
+            continue
+
+        start_time: float | None = segment["start"]
+        for word in words:
+            if start_time is None:
+                start_time = word["start"]
+            if not settings.resolve_speaker_identities_enable_split_on_words:
+                continue
+            if word["end"] - word["start"] > max_word_duration:
+                end_time = word["start"] + max_word_duration
+                if end_time > start_time:
+                    intervals.setdefault(speaker, []).append(
+                        Interval(start_time, end_time)
+                    )
+                start_time = None
+
+        if start_time is not None:
+            last = words[-1]
+            end_time = min(last["end"], last["start"] + max_word_duration)
+            if end_time > start_time:
+                intervals.setdefault(speaker, []).append(Interval(start_time, end_time))
+
+    for speaker, speaker_intervals in intervals.items():
+        intervals[speaker] = _merge_intervals(speaker_intervals)
+    return intervals
+
+
+def resolve_speaker_identities(
+    metadata: dict[str, Any],
+    transcription: Any,
+    recording_start_datetime: datetime,
+    recording_end_datetime: datetime,
+    overlap_threshold: float = settings.resolve_speaker_identities_default_overlap_threshold,  # noqa: E501
+) -> AssignmentResult:
+    """Match WhisperX speaker labels to participants.
+
+    Args:
+        metadata: User metadata with `events` and `participants`.
+        transcription: WhisperX Transcription object with a `segments` attribute.
+        recording_start_datetime: UTC datetime for t=0 reference.
+        recording_end_datetime: UTC datetime of recording end. Open speech
+            intervals are closed at this time.
+        overlap_threshold: Minimum overlap/speaker_duration to accept.
+
+    Returns:
+        AssignmentResult with per-speaker assignments and unassigned
+        speakers.
+    """
+    participant_timelines, participant_names = _build_participant_timelines(
+        metadata, recording_start_datetime, recording_end_datetime
+    )
+    speaker_timelines = _build_speaker_timelines(transcription)
+
+    logger.debug(
+        "Assignment inputs: %d participants, %d speakers\n%s\n%s\n%s",
+        len(participant_timelines),
+        len(speaker_timelines),
+        participant_timelines,
+        speaker_timelines,
+        _format_timelines_debug(
+            participant_timelines, participant_names, speaker_timelines
+        ),
+    )
+
+    result = AssignmentResult()
+
+    for speaker, speaker_intervals in speaker_timelines.items():
+        speaker_duration = _total_duration(speaker_intervals)
+        if speaker_duration == 0:
+            result.unassigned_speakers.append(speaker)
+            continue
+
+        best_pid: str | None = None
+        best_score: float = 0.0
+
+        for pid, part_intervals in participant_timelines.items():
+            overlap = _overlap_duration(speaker_intervals, part_intervals)
+            score = overlap / speaker_duration
+            if score > best_score:
+                best_score = score
+                best_pid = pid
+
+        if best_pid is not None and best_score >= overlap_threshold:
+            result.assignments.append(
+                SpeakerAssignment(
+                    speaker_label=speaker,
+                    participant_id=best_pid,
+                    participant_name=participant_names.get(best_pid, best_pid),
+                    score=best_score,
+                )
+            )
+            logger.info(
+                "Assigned %s -> %s (score=%.3f)",
+                speaker,
+                participant_names.get(best_pid, best_pid),
+                best_score,
+            )
+        else:
+            result.unassigned_speakers.append(speaker)
+            logger.info(
+                "Speaker %s unassigned (best=%.3f, threshold=%.3f)",
+                speaker,
+                best_score,
+                overlap_threshold,
+            )
+
+    return result
@@ -19,14 +19,14 @@ class TestTasks:
            headers={"Authorization": "Bearer test-api-token"},
            json={
                "owner_id": "owner-123",
-                "filename": "recording.mp4",
+                "recording_filename": "recording.mp4",
+                "metadata_filename": "metadata.json",
                "email": "user@example.com",
                "sub": "sub-123",
                "room": "room-abc",
-                "recording_date": "2026-01-01",
-                "recording_time": "10:00:00",
+                "owner_timezone": "UTC",
                "language": None,
-                "download_link": "http://example.com/file.mp4",
+                "download_link": "https://example.com/file.mp4",
            },
        )

@@ -36,16 +36,18 @@ class TestTasks:
        args = mock_apply_async.call_args.kwargs["args"]
        assert args == [
            "owner-123",  # owner_id
-            "recording.mp4",  # filename
+            "recording.mp4",  # recording_filename
+            "metadata.json",  # metadata_filename
            "user@example.com",  # email
            "sub-123",  # sub
-            1735725600.0,  # frozen time
+            1735725600.0,  # received_at
            "room-abc",  # room
-            "2026-01-01",  # recording_date
-            "10:00:00",  # recording_time
+            "UTC",  # owner_timezone
            None,  # language
-            "http://example.com/file.mp4",  # download_link
+            "https://example.com/file.mp4",  # download_link
            None,  # context_language
+            None,  # recording_start_at
+            None,  # recording_end_at
        ]

    def test_create_task_invalid_language(self, client):
@@ -0,0 +1,30 @@
+"""Unit tests for the file service."""
+
+from pathlib import Path
+
+import pytest
+
+from summary.core.file_service import get_media_duration
+
+
+@pytest.mark.parametrize(
+    "filename, duration",
+    [
+        ("audio-sample-android-chrome.webm", 2.2795),
+        ("audio-sample-android-firefox.ogg", 2.3025),
+        ("audio-sample-android.m4a", 1.38),
+        ("audio-sample-chromium.webm", 2.65),
+        ("audio-sample-firefox.ogg", 2.0865),
+        ("audio-sample-ios-browser.webm", 2.6229),
+        ("audio-sample-ios.m4a", 1.408),
+        ("audio-sample-mac-os-safari.webm", 2.3049),
+        ("video-sample-visio.mp4", 5.34059),
+    ],
+)
+def test_validate_duration_supports_all_used_file_formats(
+    filename: str, duration: float
+) -> None:
+    """Validate duration for Safari iPhone WebM files without format duration."""
+    audio_path = Path(__file__).parent.parent / "assets" / filename
+
+    assert get_media_duration(audio_path) == pytest.approx(duration, 1e-3)
@@ -0,0 +1,708 @@
+"""Tests for the speaker-to-user assignment service."""
+
+import math
+from dataclasses import dataclass, field
+from datetime import datetime
+
+from summary.core import user_assign
+from summary.core.user_assign import (
+    AssignmentResult,
+    Interval,
+    SpeakerAssignment,
+    _build_speaker_timelines,
+    _merge_intervals,
+    _overlap_duration,
+    _total_duration,
+    resolve_speaker_identities,
+)
+
+
+@dataclass
+class FakeTranscription:
+    """Mimics the OpenAI Transcription pydantic model for testing."""
+
+    segments: list = field(default_factory=list)
+
+
+RECORDING_START = datetime.fromisoformat("2026-03-17T15:30:33.000001")
+RECORDING_END = datetime.fromisoformat("2026-03-17T15:31:33.000001")
+
+METADATA_SINGLE_USER = {
+    "events": [
+        {
+            "participant_id": "da8d39ff-3b1c-4e8d-9a70-c630c9871bcb",
+            "type": "participant_connected",
+            "timestamp": "2026-03-17T15:30:33.000001",
+        },
+        {
+            "participant_id": "da8d39ff-3b1c-4e8d-9a70-c630c9871bcb",
+            "type": "speech_start",
+            "timestamp": "2026-03-17T15:30:36.039456",
+        },
+        {
+            "participant_id": "da8d39ff-3b1c-4e8d-9a70-c630c9871bcb",
+            "type": "speech_end",
+            "timestamp": "2026-03-17T15:30:36.589114",
+        },
+        {
+            "participant_id": "da8d39ff-3b1c-4e8d-9a70-c630c9871bcb",
+            "type": "speech_start",
+            "timestamp": "2026-03-17T15:30:38.887518",
+        },
+        {
+            "participant_id": "da8d39ff-3b1c-4e8d-9a70-c630c9871bcb",
+            "type": "speech_end",
+            "timestamp": "2026-03-17T15:30:39.438141",
+        },
+        {
+            "participant_id": "da8d39ff-3b1c-4e8d-9a70-c630c9871bcb",
+            "type": "participant_disconnected",
+            "timestamp": "2026-03-17T15:30:43.223255",
+        },
+    ],
+    "participants": [
+        {
+            "participantId": "da8d39ff-3b1c-4e8d-9a70-c630c9871bcb",
+            "name": "cameledev",
+        }
+    ],
+}
+
+DIARIZATION_SINGLE_SPEAKER = FakeTranscription(
+    segments=[
+        {
+            "start": 1.363,
+            "end": 3.545,
+            "text": " The stale smell.",
+            "speaker": "SPEAKER_00",
+            "words": [
+                {"word": "The", "start": 1.363, "end": 1.8},
+                {"word": "stale", "start": 1.8, "end": 2.7},
+                {"word": "smell.", "start": 2.7, "end": 3.545},
+            ],
+        },
+        {
+            "start": 4.466,
+            "end": 6.247,
+            "text": "It takes heat.",
+            "speaker": "SPEAKER_00",
+            "words": [
+                {"word": "It", "start": 4.466, "end": 4.7},
+                {"word": "takes", "start": 4.7, "end": 5.5},
+                {"word": "heat.", "start": 5.5, "end": 6.247},
+            ],
+        },
+    ],
+)
+
+USER_ID = "da8d39ff-3b1c-4e8d-9a70-c630c9871bcb"
+
+
+class TestMergeIntervals:
+    """Tests for _merge_intervals."""
+
+    def test_empty(self):
+        """Empty input returns empty list."""
+        assert _merge_intervals([]) == []
+
+    def test_no_overlap(self):
+        """Non-overlapping intervals stay separate."""
+        result = _merge_intervals([Interval(1, 2), Interval(3, 4)])
+        assert len(result) == 2
+
+    def test_overlap(self):
+        """Overlapping intervals are merged."""
+        result = _merge_intervals([Interval(1, 3), Interval(2, 4)])
+        assert len(result) == 1
+        assert result[0].start == 1
+        assert result[0].end == 4
+
+    def test_adjacent(self):
+        """Adjacent intervals are merged."""
+        result = _merge_intervals([Interval(1, 2), Interval(2, 3)])
+        assert len(result) == 1
+        assert result[0].start == 1
+        assert result[0].end == 3
+
+    def test_unsorted(self):
+        """Unsorted input is sorted before merging."""
+        result = _merge_intervals([Interval(5, 6), Interval(1, 2)])
+        assert len(result) == 2
+        assert result[0].start == 1
+
+
+class TestOverlapDuration:
+    """Tests for _overlap_duration."""
+
+    def test_no_overlap(self):
+        """Disjoint intervals have zero overlap."""
+        a = [Interval(1, 2)]
+        b = [Interval(3, 4)]
+        assert math.isclose(_overlap_duration(a, b), 0.0)
+
+    def test_full_overlap(self):
+        """Identical intervals have full overlap."""
+        a = [Interval(1, 3)]
+        b = [Interval(1, 3)]
+        assert math.isclose(_overlap_duration(a, b), 2.0)
+
+    def test_partial_overlap(self):
+        """Partially overlapping intervals."""
+        a = [Interval(1, 3)]
+        b = [Interval(2, 4)]
+        assert math.isclose(_overlap_duration(a, b), 1.0)
+
+    def test_multiple_intervals(self):
+        """Multiple intervals with a spanning interval."""
+        a = [Interval(1, 3), Interval(5, 7)]
+        b = [Interval(2, 6)]
+        assert math.isclose(_overlap_duration(a, b), 2.0)
+
+    def test_empty(self):
+        """Empty input yields zero overlap."""
+        assert math.isclose(_overlap_duration([], [Interval(1, 2)]), 0.0)
+        assert math.isclose(_overlap_duration([Interval(1, 2)], []), 0.0)
+
+
+class TestTotalDuration:
+    """Tests for _total_duration."""
+
+    def test_basic(self):
+        """Sum of durations for multiple intervals."""
+        ivs = [Interval(0, 1), Interval(2, 5)]
+        assert math.isclose(_total_duration(ivs), 4.0)
+
+    def test_empty(self):
+        """Empty input returns zero."""
+        assert math.isclose(_total_duration([]), 0.0)
+
+
+class TestBuildSpeakerTimelines:
+    """Tests for _build_speaker_timelines."""
+
+    def test_segment_without_words_falls_back_to_segment_bounds(self):
+        """Segments missing a `words` key use the segment start/end as one interval."""
+        transcription = FakeTranscription(
+            segments=[{"start": 1.5, "end": 3.5, "speaker": "SPEAKER_00"}],
+        )
+        result = _build_speaker_timelines(transcription)
+        assert result == {"SPEAKER_00": [Interval(1.5, 3.5)]}
+
+    def test_segment_with_only_none_word_timestamps_falls_back(self):
+        """If every word has None start/end, fall back to segment bounds."""
+        transcription = FakeTranscription(
+            segments=[
+                {
+                    "start": 1.0,
+                    "end": 4.0,
+                    "speaker": "SPEAKER_00",
+                    "words": [
+                        {"word": "hi", "start": None, "end": None},
+                        {"word": "there", "start": None, "end": None},
+                    ],
+                },
+            ],
+        )
+        result = _build_speaker_timelines(transcription)
+        assert result == {"SPEAKER_00": [Interval(1.0, 4.0)]}
+
+    def test_short_words_only_uses_segment_start_and_last_word_end(self):
+        """With no overly long words, the interval runs segment start to end."""
+        transcription = FakeTranscription(
+            segments=[
+                {
+                    "start": 1.0,
+                    "end": 5.0,
+                    "speaker": "SPEAKER_00",
+                    "words": [
+                        {"word": "a", "start": 1.0, "end": 1.3},
+                        {"word": "b", "start": 1.4, "end": 1.7},
+                        {"word": "c", "start": 1.8, "end": 2.1},
+                    ],
+                },
+            ],
+        )
+        result = _build_speaker_timelines(transcription)
+        # Tail: min(2.1, 1.8 + 1.0) = 2.1
+        assert result == {"SPEAKER_00": [Interval(1.0, 2.1)]}
+
+    def test_long_word_caps_interval_at_max_duration(self):
+        """A word longer than the max-word-duration cap truncates the segment."""
+        max_word_duration = (
+            user_assign.settings.resolve_speaker_identities_max_word_duration
+        )
+        transcription = FakeTranscription(
+            segments=[
+                {
+                    "start": 0.0,
+                    "end": max_word_duration + 7,
+                    "speaker": "SPEAKER_00",
+                    "words": [
+                        {
+                            "word": "pause",
+                            "start": 0.0,
+                            "end": max_word_duration + 7,
+                        },
+                    ],
+                },
+            ],
+        )
+        result = _build_speaker_timelines(transcription)
+        assert result == {"SPEAKER_00": [Interval(0.0, max_word_duration)]}
+
+    def test_long_word_in_middle_splits_segment(self):
+        """Short words around a long word produce two intervals (before-cap + after)."""
+        transcription = FakeTranscription(
+            segments=[
+                {
+                    "start": 0.0,
+                    "end": 20.0,
+                    "speaker": "SPEAKER_00",
+                    "words": [
+                        {"word": "a", "start": 0.0, "end": 0.5},
+                        {"word": "long", "start": 1.0, "end": 15.0},
+                        {"word": "z", "start": 18.0, "end": 18.4},
+                    ],
+                },
+            ],
+        )
+        result = _build_speaker_timelines(transcription)
+        # First emit: (0.0, 1.0 + 1.0). Then start_time resets, picks up at "z" (18.0).
+        # Tail: min(18.4, 18.0 + 1.0) = 18.4. So second interval is (18.0, 18.4).
+        assert result == {
+            "SPEAKER_00": [Interval(0.0, 2.0), Interval(18.0, 18.4)],
+        }
+
+    def test_tail_word_is_capped_at_max_duration(self):
+        """The trailing word's end is capped at word.start + max_word_duration."""
+        transcription = FakeTranscription(
+            segments=[
+                {
+                    "start": 0.0,
+                    "end": 50.0,
+                    "speaker": "SPEAKER_00",
+                    "words": [
+                        {"word": "a", "start": 0.0, "end": 0.4},
+                        # Last word ends inside the cap, so the cap doesn't apply.
+                        {"word": "b", "start": 1.0, "end": 1.5},
+                    ],
+                },
+            ],
+        )
+        result = _build_speaker_timelines(transcription)
+        # Tail: min(1.5, 1.0 + 1.0) = 1.5
+        assert result == {"SPEAKER_00": [Interval(0.0, 1.5)]}
+
+    def test_split_on_words_disabled_keeps_segment_as_one_interval(self, monkeypatch):
+        """With splitting disabled, long words don't split the interval."""
+        monkeypatch.setattr(
+            user_assign,
+            "settings",
+            user_assign.settings.model_copy(
+                update={"resolve_speaker_identities_enable_split_on_words": False},
+            ),
+        )
+        transcription = FakeTranscription(
+            segments=[
+                {
+                    "start": 0.0,
+                    "end": 20.0,
+                    "speaker": "SPEAKER_00",
+                    "words": [
+                        {"word": "a", "start": 0.0, "end": 0.5},
+                        {"word": "long", "start": 1.0, "end": 15.0},
+                        {"word": "z", "start": 18.0, "end": 18.4},
+                    ],
+                },
+            ],
+        )
+        result = _build_speaker_timelines(transcription)
+        # No mid-segment split; tail caps at min(18.4, 18.0 + 1.0) = 18.4.
+        assert result == {"SPEAKER_00": [Interval(0.0, 18.4)]}
+
+    def test_multiple_speakers_keep_separate_timelines(self):
+        """Segments from different speakers populate independent timeline entries."""
+        transcription = FakeTranscription(
+            segments=[
+                {
+                    "start": 0.0,
+                    "end": 1.0,
+                    "speaker": "SPEAKER_00",
+                    "words": [{"word": "hi", "start": 0.0, "end": 0.5}],
+                },
+                {
+                    "start": 2.0,
+                    "end": 3.0,
+                    "speaker": "SPEAKER_01",
+                    "words": [{"word": "yo", "start": 2.0, "end": 2.5}],
+                },
+            ],
+        )
+        result = _build_speaker_timelines(transcription)
+        assert result == {
+            "SPEAKER_00": [Interval(0.0, 0.5)],
+            "SPEAKER_01": [Interval(2.0, 2.5)],
+        }
+
+
+class TestResolveSpeakerIdentities:
+    """Tests for resolve_speaker_identities."""
+
+    def test_single_speaker_single_user(self):
+        """Single speaker assigned to single user with low threshold."""
+        result = resolve_speaker_identities(
+            METADATA_SINGLE_USER,
+            DIARIZATION_SINGLE_SPEAKER,
+            RECORDING_START,
+            RECORDING_END,
+            overlap_threshold=0.2,
+        )
+        assert len(result.assignments) == 1
+        assert result.assignments[0].speaker_label == "SPEAKER_00"
+        assert result.assignments[0].participant_id == USER_ID
+        assert result.assignments[0].participant_name == "cameledev"
+        assert result.assignments[0].score > 0
+        assert result.unassigned_speakers == []
+
+    def test_no_vad_events(self):
+        """Participant with no speech leaves speakers unassigned."""
+        metadata = {
+            "events": [
+                {
+                    "participant_id": "user-a",
+                    "type": "participant_connected",
+                    "timestamp": "2026-03-17T15:30:33.000000",
+                },
+            ],
+            "participants": [{"participantId": "user-a", "name": "Silent User"}],
+        }
+        result = resolve_speaker_identities(
+            metadata,
+            DIARIZATION_SINGLE_SPEAKER,
+            RECORDING_START,
+            RECORDING_END,
+        )
+        assert len(result.assignments) == 0
+        assert "SPEAKER_00" in result.unassigned_speakers
+
+    def test_multiple_speakers_same_user(self):
+        """Two speakers from same mic both assigned to same user."""
+        metadata = {
+            "events": [
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_start",
+                    "timestamp": "2026-03-17T15:30:35.000000",
+                },
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_end",
+                    "timestamp": "2026-03-17T15:30:50.000000",
+                },
+            ],
+            "participants": [{"participantId": "user-a", "name": "Shared Mic"}],
+        }
+        transcription = FakeTranscription(
+            segments=[
+                {"start": 1.0, "end": 3.0, "speaker": "SPEAKER_00"},
+                {"start": 5.0, "end": 7.0, "speaker": "SPEAKER_01"},
+            ],
+        )
+        result = resolve_speaker_identities(
+            metadata, transcription, RECORDING_START, RECORDING_END
+        )
+        assert len(result.assignments) == 2
+        pids = {a.participant_id for a in result.assignments}
+        assert pids == {"user-a"}
+
+    def test_two_users_two_speakers(self):
+        """Each speaker maps to correct user by VAD overlap."""
+        metadata = {
+            "events": [
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_start",
+                    "timestamp": "2026-03-17T15:30:34.000001",
+                },
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_end",
+                    "timestamp": "2026-03-17T15:30:37.000001",
+                },
+                {
+                    "participant_id": "user-b",
+                    "type": "speech_start",
+                    "timestamp": "2026-03-17T15:30:38.000001",
+                },
+                {
+                    "participant_id": "user-b",
+                    "type": "speech_end",
+                    "timestamp": "2026-03-17T15:30:41.000001",
+                },
+            ],
+            "participants": [
+                {"participantId": "user-a", "name": "Alice"},
+                {"participantId": "user-b", "name": "Bob"},
+            ],
+        }
+        transcription = FakeTranscription(
+            segments=[
+                {"start": 1.5, "end": 3.5, "speaker": "SPEAKER_00"},
+                {"start": 5.5, "end": 7.5, "speaker": "SPEAKER_01"},
+            ],
+        )
+        result = resolve_speaker_identities(
+            metadata, transcription, RECORDING_START, RECORDING_END
+        )
+        assert len(result.assignments) == 2
+        by_speaker = {a.speaker_label: a for a in result.assignments}
+        assert by_speaker["SPEAKER_00"].participant_name == "Alice"
+        assert by_speaker["SPEAKER_01"].participant_name == "Bob"
+
+    def test_overlapping_speech_two_users(self):
+        """Simultaneous speech from two users still assigns each speaker correctly."""
+        # user-a speaks from t=1s to t=6s, user-b speaks from t=3s to t=8s
+        # (3s overlap where both are speaking)
+        # SPEAKER_00 diarization covers t=1.5–5.5 (mostly user-a)
+        # SPEAKER_01 diarization covers t=4.0–7.5 (mostly user-b)
+        metadata = {
+            "events": [
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_start",
+                    "timestamp": "2026-03-17T15:30:34.000001",
+                },
+                {
+                    "participant_id": "user-b",
+                    "type": "speech_start",
+                    "timestamp": "2026-03-17T15:30:36.000001",
+                },
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_end",
+                    "timestamp": "2026-03-17T15:30:39.000001",
+                },
+                {
+                    "participant_id": "user-b",
+                    "type": "speech_end",
+                    "timestamp": "2026-03-17T15:30:41.000001",
+                },
+            ],
+            "participants": [
+                {"participantId": "user-a", "name": "Alice"},
+                {"participantId": "user-b", "name": "Bob"},
+            ],
+        }
+        transcription = FakeTranscription(
+            segments=[
+                {"start": 1.5, "end": 5.5, "speaker": "SPEAKER_00"},
+                {"start": 4.0, "end": 7.5, "speaker": "SPEAKER_01"},
+            ],
+        )
+        result = resolve_speaker_identities(
+            metadata,
+            transcription,
+            RECORDING_START,
+            RECORDING_END,
+            overlap_threshold=0.3,
+        )
+        assert len(result.assignments) == 2
+        by_speaker = {a.speaker_label: a for a in result.assignments}
+        assert by_speaker["SPEAKER_00"].participant_name == "Alice"
+        assert by_speaker["SPEAKER_01"].participant_name == "Bob"
+        assert result.unassigned_speakers == []
+
+    def test_below_threshold(self):
+        """Speaker with minimal overlap stays unassigned."""
+        metadata = {
+            "events": [
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_start",
+                    "timestamp": "2026-03-17T15:30:34.000001",
+                },
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_end",
+                    "timestamp": "2026-03-17T15:30:35.169950",
+                },
+            ],
+            "participants": [{"participantId": "user-a", "name": "Brief User"}],
+        }
+        transcription = FakeTranscription(
+            segments=[
+                {"start": 1.0, "end": 10.0, "speaker": "SPEAKER_00"},
+            ],
+        )
+        result = resolve_speaker_identities(
+            metadata,
+            transcription,
+            RECORDING_START,
+            RECORDING_END,
+            overlap_threshold=0.5,
+        )
+        assert len(result.assignments) == 0
+        assert "SPEAKER_00" in result.unassigned_speakers
+
+    def test_events_before_recording_start_clamped(self):
+        """Speech events before recording start are clamped to t=0."""
+        metadata = {
+            "events": [
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_start",
+                    "timestamp": "2026-03-17T15:30:31.000001",  # before RECORDING_START
+                },
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_end",
+                    "timestamp": "2026-03-17T15:30:36.000001",  # after RECORDING_START
+                },
+            ],
+            "participants": [{"participantId": "user-a", "name": "Early User"}],
+        }
+        transcription = FakeTranscription(
+            segments=[
+                {"start": 0.0, "end": 3.0, "speaker": "SPEAKER_00"},
+            ],
+        )
+        result = resolve_speaker_identities(
+            metadata, transcription, RECORDING_START, RECORDING_END
+        )
+        assert len(result.assignments) == 1
+        assert result.assignments[0].participant_name == "Early User"
+
+    def test_empty_diarization(self):
+        """No segments produces empty result."""
+        result = resolve_speaker_identities(
+            METADATA_SINGLE_USER,
+            FakeTranscription(segments=[]),
+            RECORDING_START,
+            RECORDING_END,
+        )
+        assert result == AssignmentResult()
+
+    def test_segment_without_speaker_ignored(self):
+        """Segments missing speaker key are skipped."""
+        transcription = FakeTranscription(
+            segments=[
+                {"start": 1.0, "end": 3.0, "text": "no speaker"},
+            ],
+        )
+        result = resolve_speaker_identities(
+            METADATA_SINGLE_USER, transcription, RECORDING_START, RECORDING_END
+        )
+        assert result == AssignmentResult()
+
+    def test_unclosed_speech_closed_at_recording_end(self):
+        """Open speech_start without speech_end is closed at recording end."""
+        recording_end = datetime.fromisoformat("2026-03-17T15:30:43.000001")
+        metadata = {
+            "events": [
+                {
+                    "participant_id": "user-a",
+                    "type": "speech_start",
+                    "timestamp": "2026-03-17T15:30:35.000001",
+                },
+                # No speech_end — participant kept speaking until recording stopped
+            ],
+            "participants": [{"participantId": "user-a", "name": "Still Talking"}],
+        }
+        transcription = FakeTranscription(
+            segments=[
+                {"start": 2.0, "end": 9.0, "speaker": "SPEAKER_00"},
+            ],
+        )
+        result = resolve_speaker_identities(
+            metadata,
+            transcription,
+            RECORDING_START,
+            recording_end,
+            overlap_threshold=0.5,
+        )
+        assert len(result.assignments) == 1
+        assert result.assignments[0].participant_name == "Still Talking"
+        assert result.unassigned_speakers == []
+
+
+class TestApply:
+    """Tests for AssignmentResult.apply_to."""
+
+    def test_replaces_speakers_in_segments_and_words(self):
+        """Speaker labels replaced in segments, words, and word_segments."""
+        diarization = {
+            "segments": [
+                {
+                    "start": 1.0,
+                    "end": 3.0,
+                    "text": "Hello",
+                    "speaker": "SPEAKER_00",
+                    "words": [
+                        {
+                            "word": "Hello",
+                            "start": 1.0,
+                            "end": 1.5,
+                            "speaker": "SPEAKER_00",
+                        },
+                    ],
+                },
+                {
+                    "start": 4.0,
+                    "end": 6.0,
+                    "text": "World",
+                    "speaker": "SPEAKER_01",
+                    "words": [
+                        {
+                            "word": "World",
+                            "start": 4.0,
+                            "end": 4.5,
+                            "speaker": "SPEAKER_01",
+                        },
+                    ],
+                },
+            ],
+            "word_segments": [
+                {"word": "Hello", "start": 1.0, "end": 1.5, "speaker": "SPEAKER_00"},
+                {"word": "World", "start": 4.0, "end": 4.5, "speaker": "SPEAKER_01"},
+            ],
+        }
+        assignment = AssignmentResult(
+            assignments=[
+                SpeakerAssignment("SPEAKER_00", "id-a", "Alice", 0.9),
+                SpeakerAssignment("SPEAKER_01", "id-b", "Bob", 0.8),
+            ],
+        )
+        result = assignment.apply_to(diarization)
+
+        assert result["segments"][0]["speaker"] == "Alice"
+        assert result["segments"][0]["words"][0]["speaker"] == "Alice"
+        assert result["segments"][1]["speaker"] == "Bob"
+        assert result["word_segments"][0]["speaker"] == "Alice"
+        assert result["word_segments"][1]["speaker"] == "Bob"
+
+    def test_unassigned_speakers_unchanged(self):
+        """Unassigned speaker labels are left as-is."""
+        diarization = {
+            "segments": [
+                {"start": 1.0, "end": 3.0, "speaker": "SPEAKER_02"},
+            ],
+        }
+        assignment = AssignmentResult(
+            assignments=[
+                SpeakerAssignment("SPEAKER_00", "id-a", "Alice", 0.9),
+            ],
+            unassigned_speakers=["SPEAKER_02"],
+        )
+        result = assignment.apply_to(diarization)
+        assert result["segments"][0]["speaker"] == "SPEAKER_02"
+
+    def test_preserves_extra_keys(self):
+        """Non-segment keys in diarization are preserved."""
+        diarization = {
+            "segments": [],
+            "language": "en",
+            "custom_field": 42,
+        }
+        result = AssignmentResult().apply_to(diarization)
+        assert result["language"] == "en"
+        assert result["custom_field"] == 42
Author	SHA1	Message	Date
lebaudantoine	7fd8e3a242	wip align naming	2026-05-13 17:14:38 +02:00
lebaudantoine	17f5bb50cb	wip fix suggestion	2026-05-13 17:12:29 +02:00
leo	96f97ed2d0	✨(summary) improve speaker assignment Speaker-to-participant assignment relie on WhisperX word timings, but incorrect word durations in the output can lead to inaccurate overlap scoring and wrong user attribution. Add a custom heuristic to trim overly long word durations before computing assignments.	2026-05-12 16:58:07 +02:00
lebaudantoine	02d16cb55c	⬆️(addons) update dependencies	2026-05-12 16:26:16 +02:00
lebaudantoine	7268ff6777	⬆️(mail) update dependencies	2026-05-12 16:26:16 +02:00
lebaudantoine	cca5bc2186	⬆️(frontend) update dependencies	2026-05-12 16:26:16 +02:00
leo	ec67a12fe4	✨(agents) use uv for dependency management Change from pip to uv for dependancy management in src/agents.	2026-05-12 13:47:19 +02:00
leo	05f32d008a	⬆️ (dependencies) Bump urllib3 from 2.6.3 to 2.7.0 [SECURITY] Fix CVE-2026-44431 and CVE-2026-44432.	2026-05-12 11:23:00 +02:00
UGilfoyle	964b3cd452	🐛(backend) add link to "Open" text in recording email Added a hyperlink to the "Open" text in step 1 of the recording notification email instructions. Previously, "Open" was plain text and users could only access their recording via the button below. Now the text itself is a clickable link, improving accessibility for email clients that may not render the button properly. Updated MJML source template and all 4 locale files (en, fr, de, nl).	2026-05-11 23:04:27 +02:00
Florent Chehab	c7ca5a621f	🐛(ci) install ffmpeg for summary tests Add ffmpeg for summary tests	2026-05-11 23:00:55 +02:00
Florent Chehab	90ebe231ef	🐛(summary) complete webm support When duration is not reported in the files metadata, we directly infer the duration from the audio packets. This prevents errors on webm files. Very simple audio & video test files have been added that cover relevant usecases to prevent regressions.	2026-05-11 23:00:54 +02:00
soyouzpanda	04f2a9ebdc	⬆️(mail) fix dependencies not having resolved or integrity field Update dependencies to the latest minor versions fixed that by re-resolving the fields. This is needed for packaging as many distribution retrieve node modules into the npm cache and then tries to install node modules into the project without any internet connection. Since there is no resolved/integrity field, it fails to get packages from the cache.	2026-05-11 12:45:02 +02:00
renovate[bot]	6a8eb79b41	⬆️(dependencies) update django to v5.2.14 [SECURITY]	2026-05-11 12:03:18 +02:00
leo	bc35046b3a	🩹(summary) fix bug in assign_user Fix bug in speaker assignment which occurs when LIVEKIT_VERIFY_SSL is True.	2026-05-07 18:17:15 +02:00
leo	1612d8b2d4	✨(audio) assign users to diarization speaker results using VAD Introduce a new user assignment mechanism to for more friendly output than the current (SPEAKER_0, SPEAKER_1, ...). Use the VAD metadata to compare speech intervals with those returned by WhisperX. User with the highest overlap score above a defined threshold is assigned to each segment. This method allows for multi-speaker scenarios for a single account.	2026-05-07 12:45:00 +02:00
lebaudantoine	f8937fc0a1	♻️(frontend) improve and simplify accessibility font override logic Fix compatibility issues with the DINUM frontend image, which overrides the default `font-sans` value. Simplify the implementation by having the JavaScript layer only toggle well-scoped CSS classes responsible for accessibility font overrides. This makes the behavior more predictable and restoring default styles straightforward. Also clarify the intent of the hook by making its accessibility purpose explicit and moving its usage to the App component, where it better fits the application lifecycle.	2026-05-07 11:20:15 +02:00
Cyril	97b5e3e65c	✨(frontend) add font selector in accessibility settings Dropdown with description and FR/EN/NL translations.	2026-05-07 11:20:15 +02:00
Cyril	b917d82f7e	✨(frontend) apply font preference to app layout Hook, CSS variable and LiveKit integration for custom fonts.	2026-05-07 11:20:15 +02:00
Cyril	82d146cdf5	✨(frontend) install accessibility font packages Lexend, Atkinson Hyperlegible Next and OpenDyslexic via fontsource.	2026-05-07 11:20:15 +02:00
Cyril	cbfeea0a4e	✨(frontend) add uiFont preference to accessibility store Add UiFont type with four options and Extend AccessibilityState.	2026-05-07 11:20:15 +02:00
leo	a695758da4	♻️(summary) refactor tasks signature and make transcription tz-aware The tasks endpoint used non-timezone-aware date and time values and split them into separate variables, which is unconventional. Refactor the implementation to use timezone-aware datetime objects and align transcription formatting with the user-declared timezone. Update the source of truth for recording start time to FileInfo.started_at for improved precision. Adjust the task signature in preparation for upcoming user assignment work, which will require `started_at`, `ended_at`, and `metadata_filename`.	2026-05-06 18:33:03 +02:00
Damien Laine	4c5b6de8f3	✨(backend) make LiveKit Egress recording encoding configurable Expose RECORDING_ENCODING_* settings to override the default LiveKit Egress preset (H264_720P_30). When RECORDING_ENCODING_ENABLED is True, the provided width/height/framerate/bitrate/keyframe values are passed as advanced EncodingOptions. Lowering framerate and bitrate reduces recording file size and egress worker CPU load. Disabled by default, preserving current behaviour.	2026-05-05 18:26:49 +02:00
Florent Chehab	cf4e347589	✨(helm) add support multiple transcribe worker / endpoint Udate the helm chart to support multiple transcribe worker in the summary service. This is useful when using multiple WhisperX instances to have one deployment for each endpoint. This enables some kind of horizontal scaling (we still keep one call per WhisperX endpoint but can have multiple WhisperX endpoints)	2026-05-05 09:16:50 +02:00
tuanaiseo	fc260b2686	🔒️(frontend) room ids are generated with non-cryptographic rand Room identifiers are created with `Math.random()`, which is predictable and not suitable for security-sensitive identifiers. Predictable room IDs increase the risk of room enumeration and unauthorized access attempts, especially when IDs are part of join URLs. Affected files: generateRoomId.ts Signed-off-by: tuanaiseo <221258316+tuanaiseo@users.noreply.github.com>	2026-05-04 23:47:47 +02:00
lebaudantoine	cd7799997e	🧑‍💻(bin) update release tooling to support uv-based deps management Following the switch from pip to uv, prepare the release workflow to automatically run `uv lock` on backend and keep dependencies up to date.	2026-05-04 22:36:07 +02:00
EpsilonFO	a2bccf4f4f	🐛(backend) make start-recording atomic and fault-tolerant Wrap Recording and RecordingAccess creation in a single transaction so a partial failure does not leave orphan rows, and return 409 instead of 500 when a recording is already in progress for the room. When the worker fails to start, transition the Recording to FAILED_TO_START so the unique partial constraint on (room, status) no longer blocks future recording attempts on the same room.	2026-05-04 22:15:14 +02:00
Sanjay Santhanam	6830250f2c	♻️(frontend) standardize role terminology across localizations Fixes #1126 - Inconsistent role terminology in localization files. Standardize on 'host' as the primary role term across en, de, and nl locales, replacing mixed usage of 'administrator', 'organizer', 'admin', 'Organisator:in', 'Organisierende', and 'organisator'.	2026-05-04 18:44:20 +02:00
leo	0c0ce87947	🔒️(backend) validate Room configuration with Pydantic schema Room.configuration accepted arbitrary JSON without validation, allowing unsafe or malformed payloads to be stored and creating a security risk. Define a Pydantic schema to enforce structure and constraints, and add validation at the serializer level to reject invalid inputs.	2026-05-04 18:10:44 +02:00
renovate[bot]	597eba6e8a	⬆️(dependencies) update postcss to v8.5.10 [SECURITY]	2026-05-04 16:48:46 +02:00
renovate[bot]	47dbc271ba	⬆️(dependencies) update webpack-dev-server to v5.2.1 [SECURITY]	2026-05-04 15:34:30 +02:00
renovate[bot]	c3adcc8ff3	⬆️(dependencies) update pytest to v9.0.3 [SECURITY]	2026-05-04 14:24:53 +02:00