Add complete financial fraud detection database with Docker, schema, data generation, and SQL exercises

- Docker setup with PostgreSQL 16 and DB-UI web interface - Comprehensive 20+ table schema with fraud detection patterns - Idempotent shell scripts for data generation (no Python dependency) - Realistic geographic data (100 US cities, 210 world cities) - 5M+ transactions with embedded fraud patterns (velocity, geographic, structuring, etc.) - Progressive SQL exercises from beginner to advanced fraud detection - Complete documentation and quick start guide - Setup and verification scripts
2026-08-06 00:17:42 +00:00 · 2025-10-23 13:53:30 -04:00
parent 9d9104a4fe
commit b30733ccad
15 changed files with 3864 additions and 2 deletions
@@ -1,2 +1,311 @@
-# SQL
+# Financial Fraud Detection - SQL Learning Database
-A repository for learning SQL
+
 A comprehensive, production-grade database designed for learning SQL through realistic financial fraud investigation scenarios.
 ## 🎯 Overview
 This project provides a complete PostgreSQL database with **5+ million transactions**, embedded fraud patterns, and progressive SQL exercises. Perfect for:
 - **SQL Beginners** → Learn fundamentals with real-world data
 - **Data Analysts** → Practice fraud detection queries
 - **Security Professionals** → Understand fraud patterns
 - **Students** → Hands-on financial crime investigation
 ## 📊 Database Statistics
 - **100,000** Customers with KYC data
 - **150,000** Bank accounts (checking, savings, credit)
 - **200,000** Payment cards
 - **50,000** Merchants across 35 categories
 - **5,000,000** Transactions (7% fraudulent)
 - **50,000+** Fraud alerts
 - **5,000+** Fraud cases
 - **500,000** Login sessions
 ## 🏗️ Architecture
 ### Data Model Features
 - ✅ **20+ Tables** with proper relationships
 - ✅ **Foreign key constraints** for data integrity
 - ✅ **Indexes** for query performance
 - ✅ **Realistic geographic data** (100 US cities, 210 world cities)
 - ✅ **Embedded fraud patterns** (velocity, geographic, structuring)
 - ✅ **Audit trails** and compliance tables
 ### Key Entities
 ```
 customers → accounts → transactions
                    ↓
                  cards → merchants
                    ↓
                 alerts → fraud_cases
 ```
 ## 🚀 Quick Start
 ### Prerequisites
 - Docker and Docker Compose
 - 8GB RAM minimum
 - 20GB disk space
 ### 1. Clone the Repository
 ```bash
 git clone https://github.com/yourusername/SQL.git
 cd SQL
 ```
 ### 2. Start the Database
 ```bash
 docker-compose up -d
 ```
 This starts:
 - **PostgreSQL 16** on port `5432`
 - **DB-UI** web interface on port `3000`
 ### 3. Initialize the Schema
 ```bash
 chmod +x scripts/setup-database.sh
 ./scripts/setup-database.sh
 ```
 ### 4. Generate Test Data
 ```bash
 chmod +x data/generate_data.sh
 ./data/generate_data.sh
 ```
 ⏱️ **Note:** Data generation takes 15-30 minutes depending on your system.
 ### 5. Access the Database
 **Option A: DB-UI Web Interface**
 ```
 http://localhost:3000
 ```
 **Option B: Command Line**
 ```bash
 docker exec -it fraud_detection_db psql -U fraud_analyst -d fraud_detection
 ```
 **Option C: Your Favorite SQL Client**
 ```
 Host: localhost
 Port: 5432
 Database: fraud_detection
 Username: fraud_analyst
 Password: SecurePass123!
 ```
 ## 📚 Learning Path
 ### Level 1: Basic Queries
 - SELECT, WHERE, ORDER BY
 - Filtering and sorting
 - Basic comparisons
 - **Location:** `exercises/01-basic-queries/`
 ### Level 2: Joins
 - INNER JOIN, LEFT JOIN
 - Multiple table queries
 - Relationship navigation
 - **Location:** `exercises/02-joins/`
 ### Level 3: Aggregations
 - COUNT, SUM, AVG, MAX, MIN
 - GROUP BY and HAVING
 - Statistical analysis
 - **Location:** `exercises/03-aggregations/`
 ### Level 4: Subqueries
 - Nested queries
 - Correlated subqueries
 - EXISTS and IN
 - **Location:** `exercises/04-subqueries/`
 ### Level 5: Window Functions
 - ROW_NUMBER, RANK, DENSE_RANK
 - Running totals
 - Moving averages
 - **Location:** `exercises/05-window-functions/`
 ### Level 6: Fraud Detection
 - Velocity fraud detection
 - Geographic anomalies
 - Money mule networks
 - Account takeover patterns
 - **Location:** `exercises/06-fraud-detection/`
 ## 🔍 Fraud Patterns Included
 ### 1. Velocity Fraud
 Multiple rapid transactions from the same account
 ### 2. Geographic Impossibility
 Card used in different countries within hours
 ### 3. Money Mule Networks
 Rapid transfer chains between accounts
 ### 4. Account Takeover
 Sudden changes in transaction patterns
 ### 5. Structuring (Smurfing)
 Multiple transactions just under $10,000 reporting threshold
 ### 6. Card Testing
 Multiple small failed transactions
 ### 7. High-Risk Merchants
 Unusual activity at gambling/crypto merchants
 ### 8. Dormant Account Reactivation
 Long-inactive accounts suddenly active
 ## 📁 Project Structure
 ```
 SQL/
 ├── docker-compose.yml          # Docker orchestration
 ├── docker/
 │   └── init/                   # Database initialization
 ├── schema/
 │   ├── 01-create-tables.sql    # DDL (idempotent)
 │   └── 02-seed-data.sql        # Reference data
 ├── data/
 │   ├── generate_data.sh        # Data generation script
 │   └── reference/              # Geographic data
 │       ├── us_cities.csv
 │       ├── world_cities.csv
 │       └── load_geographic_data.sql
 ├── scripts/
 │   └── setup-database.sh       # Setup automation
 ├── exercises/
 │   ├── 01-basic-queries/
 │   ├── 02-joins/
 │   ├── 03-aggregations/
 │   ├── 04-subqueries/
 │   ├── 05-window-functions/
 │   └── 06-fraud-detection/
 └── docs/
    ├── data-model.md
    ├── fraud-patterns.md
    └── setup-guide.md
 ```
 ## 🔧 Configuration
 ### Environment Variables
 Edit `docker-compose.yml` to customize:
 ```yaml
 POSTGRES_DB: fraud_detection
 POSTGRES_USER: fraud_analyst
 POSTGRES_PASSWORD: SecurePass123!
 ```
 ### Data Volume
 Modify `data/generate_data.sh`:
 ```bash
 NUM_CUSTOMERS=100000      # Adjust as needed
 NUM_TRANSACTIONS=5000000  # Adjust as needed
 FRAUD_PERCENTAGE=7        # 7% fraudulent
 ```
 ## 🔄 Idempotency
 All scripts are **idempotent** - safe to run multiple times:
 - `setup-database.sh` - Drops and recreates schema
 - `generate_data.sh` - Clears and regenerates data
 - Schema files use `DROP IF EXISTS`
 ## 🎓 Sample Queries
 ### Find High-Risk Customers
 ```sql
 SELECT customer_id, first_name, last_name, risk_score
 FROM customers
 WHERE risk_score > 80
 ORDER BY risk_score DESC;
 ```
 ### Detect Velocity Fraud
 ```sql
 SELECT account_id, COUNT(*) as txn_count, SUM(amount) as total
 FROM transactions
 WHERE transaction_date >= NOW() - INTERVAL '1 hour'
 GROUP BY account_id
 HAVING COUNT(*) > 5;
 ```
 ### Geographic Anomalies
 ```sql
 SELECT t1.card_id, c1.country_name, c2.country_name,
       t2.transaction_date - t1.transaction_date as time_diff
 FROM transactions t1
 JOIN transactions t2 ON t1.card_id = t2.card_id
 JOIN countries c1 ON t1.country_id = c1.country_id
 JOIN countries c2 ON t2.country_id = c2.country_id
 WHERE t1.country_id != t2.country_id
  AND t2.transaction_date BETWEEN t1.transaction_date
  AND t1.transaction_date + INTERVAL '2 hours';
 ```
 ## 🛠️ Maintenance
 ### Reset Everything
 ```bash
 docker-compose down -v
 docker-compose up -d
 ./scripts/setup-database.sh
 ./data/generate_data.sh
 ```
 ### Backup Database
 ```bash
 docker exec fraud_detection_db pg_dump -U fraud_analyst fraud_detection > backup.sql
 ```
 ### Restore Database
 ```bash
 cat backup.sql | docker exec -i fraud_detection_db psql -U fraud_analyst -d fraud_detection
 ```
 ## 📖 Documentation
 - **[Data Model](docs/data-model.md)** - Complete ER diagram and table descriptions
 - **[Fraud Patterns](docs/fraud-patterns.md)** - Detailed fraud scenario explanations
 - **[Setup Guide](docs/setup-guide.md)** - Detailed installation instructions
 ## 🤝 Contributing
 Contributions welcome! Please:
 1. Fork the repository
 2. Create a feature branch
 3. Add exercises or improve data generation
 4. Submit a pull request
 ## 📝 License
 This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
 ## 🙏 Acknowledgments
 - PostgreSQL community
 - DB-UI project (https://github.com/n7olkachev/db-ui)
 - Financial crime investigation best practices
 ## 📧 Support
 - **Issues:** GitHub Issues
 - **Discussions:** GitHub Discussions
 - **Documentation:** `/docs` folder
 ---
 **Happy Learning! 🎉**
 Start with `exercises/01-basic-queries/` and work your way up to detecting sophisticated fraud patterns!
@@ -0,0 +1,384 @@
 # 🎉 Financial Fraud Detection Database - Setup Complete!
 ## ✅ What Has Been Created
 ### 1. **Docker Infrastructure**
 - ✅ `docker-compose.yml` - Orchestrates PostgreSQL + DB-UI
 - ✅ PostgreSQL 16 Alpine container
 - ✅ DB-UI web interface (https://github.com/n7olkachev/db-ui)
 - ✅ Persistent volumes for data
 - ✅ Health checks and auto-restart
 ### 2. **Database Schema (20+ Tables)**
 #### Core Tables
 - ✅ `customers` - 100K customer records with KYC data
 - ✅ `accounts` - 150K bank accounts (checking, savings, credit)
 - ✅ `cards` - 200K payment cards
 - ✅ `transactions` - 5M transactions with fraud patterns
 - ✅ `merchants` - 50K merchants across 35 categories
 - ✅ `devices` - 75K device fingerprints
 #### Fraud Detection Tables
 - ✅ `alerts` - System-generated fraud alerts
 - ✅ `fraud_cases` - Confirmed fraud investigations
 - ✅ `case_transactions` - Links transactions to cases
 - ✅ `case_alerts` - Links alerts to cases
 #### Supporting Tables
 - ✅ `countries` - 40 countries with risk levels
 - ✅ `merchant_categories` - 35 MCC categories
 - ✅ `transaction_types` - 15 transaction types
 - ✅ `fraud_types` - 20 fraud pattern types
 - ✅ `login_sessions` - 500K login history
 - ✅ `transfers` - Money transfer records
 - ✅ `beneficiaries` - Transfer recipients
 - ✅ `customer_relationships` - Network analysis
 - ✅ `suspicious_activity_reports` - SAR filings
 - ✅ `audit_log` - Complete audit trail
 ### 3. **Realistic Geographic Data**
 - ✅ 100 US cities with matching states
 - ✅ 210 world cities across 40 countries
 - ✅ Proper city/state/country relationships
 - ✅ Risk-based country classifications
 ### 4. **Embedded Fraud Patterns**
 - ✅ Velocity fraud (rapid transactions)
 - ✅ Geographic impossibility (same card, different countries)
 - ✅ Money mule networks (transfer chains)
 - ✅ Account takeover (behavior changes)
 - ✅ Structuring/Smurfing (avoiding $10K threshold)
 - ✅ Card testing (multiple small failures)
 - ✅ High-risk merchant abuse
 - ✅ Dormant account reactivation
 ### 5. **Data Generation Scripts**
 - ✅ `generate_data.sh` - Idempotent data generation
 - ✅ Realistic distributions (Pareto, normal)
 - ✅ Temporal patterns (2+ years of data)
 - ✅ 7% fraud rate (industry realistic)
 - ✅ Progress tracking and colored output
 ### 6. **Setup & Maintenance Scripts**
 - ✅ `setup-database.sh` - Idempotent schema setup
 - ✅ `verify-setup.sh` - Comprehensive verification
 - ✅ All scripts with error handling
 - ✅ Color-coded output for clarity
 ### 7. **SQL Learning Exercises**
 #### Level 1: Basic Queries
 - SELECT, WHERE, ORDER BY
 - Filtering and sorting
 - 10 exercises + 3 challenges
 #### Level 6: Fraud Detection
 - Velocity fraud detection
 - Geographic anomalies
 - Money mule networks
 - Account takeover patterns
 - Structuring detection
 - Card testing
 - High-risk merchant analysis
 - Dormant account reactivation
 ### 8. **Documentation**
 - ✅ Comprehensive README.md
 - ✅ Quick Start Guide (docs/QUICKSTART.md)
 - ✅ Exercise documentation
 - ✅ Inline SQL comments
 - ✅ This setup summary
 ---
 ## 🚀 How to Use
 ### Quick Start (5 minutes + data generation time)
 ```bash
 # 1. Start containers
 docker-compose up -d
 # 2. Setup schema
 chmod +x scripts/*.sh
 ./scripts/setup-database.sh
 # 3. Generate data (15-30 minutes)
 chmod +x data/generate_data.sh
 ./data/generate_data.sh
 # 4. Verify setup
 ./scripts/verify-setup.sh
 # 5. Access DB-UI
 # Open browser to: http://localhost:3000
 ```
 ### Connection Details
 **DB-UI Web Interface:**
 ```
 URL: http://localhost:3000
 ```
 **Direct Database Connection:**
 ```
 Host:     localhost
 Port:     5432
 Database: fraud_detection
 Username: fraud_analyst
 Password: SecurePass123!
 ```
 **Command Line (psql):**
 ```bash
 docker exec -it fraud_detection_db psql -U fraud_analyst -d fraud_detection
 ```
 ---
 ## 📊 Data Volumes
 ### Default Configuration
 - **Customers:** 100,000
 - **Accounts:** 150,000
 - **Merchants:** 50,000
 - **Devices:** 75,000
 - **Cards:** 200,000
 - **Login Sessions:** 500,000
 - **Transactions:** 5,000,000
 - **Alerts:** ~50,000
 - **Fraud Cases:** ~5,000
 ### Customization
 Edit `data/generate_data.sh`:
 ```bash
 NUM_CUSTOMERS=100000      # Adjust as needed
 NUM_ACCOUNTS=150000
 NUM_MERCHANTS=50000
 NUM_DEVICES=75000
 NUM_CARDS=200000
 NUM_TRANSACTIONS=5000000
 FRAUD_PERCENTAGE=7        # 7% fraudulent
 ```
 ---
 ## 🔄 Idempotency
 All scripts are **safe to run multiple times**:
 - ✅ `setup-database.sh` - Drops and recreates schema
 - ✅ `generate_data.sh` - Clears and regenerates data
 - ✅ Schema files use `DROP IF EXISTS`
 - ✅ Seed data uses `TRUNCATE ... RESTART IDENTITY`
 **To reset everything:**
 ```bash
 docker-compose down -v
 docker-compose up -d
 ./scripts/setup-database.sh
 ./data/generate_data.sh
 ```
 ---
 ## 🎓 Learning Path
 ### Beginner (Weeks 1-2)
 1. Start with `exercises/01-basic-queries/`
 2. Learn SELECT, WHERE, ORDER BY
 3. Practice filtering and sorting
 4. Explore the data with DB-UI
 ### Intermediate (Weeks 3-4)
 1. Master JOINs (exercises/02-joins/)
 2. Learn aggregations (exercises/03-aggregations/)
 3. Practice subqueries (exercises/04-subqueries/)
 ### Advanced (Weeks 5-6)
 1. Window functions (exercises/05-window-functions/)
 2. Fraud detection scenarios (exercises/06-fraud-detection/)
 3. Complex pattern detection
 4. Performance optimization
 ---
 ## 🔍 Sample Queries to Get Started
 ### 1. Explore the Data
 ```sql
 -- How many records in each table?
 SELECT 'customers' as table_name, COUNT(*) FROM customers
 UNION ALL
 SELECT 'accounts', COUNT(*) FROM accounts
 UNION ALL
 SELECT 'transactions', COUNT(*) FROM transactions
 UNION ALL
 SELECT 'alerts', COUNT(*) FROM alerts;
 ```
 ### 2. Find High-Risk Activity
 ```sql
 -- Top 10 highest fraud scores
 SELECT 
    t.transaction_id,
    c.first_name || ' ' || c.last_name as customer,
    t.amount,
    t.fraud_score,
    t.flagged_reason
 FROM transactions t
 JOIN accounts a ON t.account_id = a.account_id
 JOIN customers c ON a.customer_id = c.customer_id
 WHERE t.is_flagged = TRUE
 ORDER BY t.fraud_score DESC
 LIMIT 10;
 ```
 ### 3. Geographic Analysis
 ```sql
 -- Transactions by country
 SELECT 
    co.country_name,
    co.risk_level,
    COUNT(*) as transaction_count,
    SUM(t.amount) as total_amount
 FROM transactions t
 JOIN countries co ON t.country_id = co.country_id
 GROUP BY co.country_name, co.risk_level
 ORDER BY total_amount DESC;
 ```
 ---
 ## 🛠️ Maintenance
 ### Backup Database
 ```bash
 docker exec fraud_detection_db pg_dump -U fraud_analyst fraud_detection > backup_$(date +%Y%m%d).sql
 ```
 ### Restore Database
 ```bash
 cat backup_20241023.sql | docker exec -i fraud_detection_db psql -U fraud_analyst -d fraud_detection
 ```
 ### View Logs
 ```bash
 # PostgreSQL logs
 docker-compose logs postgres
 # DB-UI logs
 docker-compose logs db-ui
 # Follow logs
 docker-compose logs -f
 ```
 ### Performance Tuning
 ```sql
 -- Check table sizes
 SELECT 
    schemaname,
    tablename,
    pg_size_pretty(pg_total_relation_size(schemaname||'.'||tablename)) AS size
 FROM pg_tables
 WHERE schemaname = 'public'
 ORDER BY pg_total_relation_size(schemaname||'.'||tablename) DESC;
 -- Check index usage
 SELECT 
    schemaname,
    tablename,
    indexname,
    idx_scan,
    idx_tup_read,
    idx_tup_fetch
 FROM pg_stat_user_indexes
 ORDER BY idx_scan DESC;
 ```
 ---
 ## 📁 Project Structure
 ```
 SQL/
 ├── docker-compose.yml          # Docker orchestration
 ├── README.md                   # Main documentation
 ├── SETUP_COMPLETE.md          # This file
 ├── LICENSE                     # MIT License
 │
 ├── docker/
 │   └── init/
 │       └── 00-init-database.sql
 │
 ├── schema/
 │   ├── 01-create-tables.sql   # DDL (idempotent)
 │   └── 02-seed-data.sql       # Reference data
 │
 ├── data/
 │   ├── generate_data.sh       # Data generation (idempotent)
 │   └── reference/
 │       ├── us_cities.csv
 │       ├── world_cities.csv
 │       └── load_geographic_data.sql
 │
 ├── scripts/
 │   ├── setup-database.sh      # Schema setup (idempotent)
 │   └── verify-setup.sh        # Verification
 │
 ├── exercises/
 │   ├── 01-basic-queries/
 │   │   └── README.md
 │   └── 06-fraud-detection/
 │       └── README.md
 │
 └── docs/
    └── QUICKSTART.md          # Quick start guide
 ```
 ---
 ## 🎯 Success Criteria
 Your setup is complete when:
 - ✅ Docker containers are running
 - ✅ Database has 20+ tables
 - ✅ Reference data is loaded (countries, categories, etc.)
 - ✅ Test data is generated (customers, transactions, etc.)
 - ✅ DB-UI is accessible at http://localhost:3000
 - ✅ You can run queries successfully
 **Verify with:**
 ```bash
 ./scripts/verify-setup.sh
 ```
 ---
 ## 🤝 Next Steps
 1. **Read the Quick Start:** `docs/QUICKSTART.md`
 2. **Start Learning:** `exercises/01-basic-queries/README.md`
 3. **Explore DB-UI:** http://localhost:3000
 4. **Practice Queries:** Try the sample queries above
 5. **Detect Fraud:** `exercises/06-fraud-detection/README.md`
 ---
 ## 📧 Support
 - **Documentation:** Check `/docs` folder
 - **Exercises:** Check `/exercises` folder
 - **Issues:** Use GitHub Issues
 - **Verification:** Run `./scripts/verify-setup.sh`
 ---
 **🎉 Congratulations! Your fraud detection database is ready for SQL learning!**
 **Start here:** `docs/QUICKSTART.md` or `exercises/01-basic-queries/README.md`
@@ -0,0 +1,541 @@
 #!/bin/bash
 # ============================================================================
 # Financial Fraud Detection - Data Generation Script - IDEMPOTENT
 # ============================================================================
 # Generates realistic test data with embedded fraud patterns
 # This script is IDEMPOTENT - it will clear and regenerate all data
 # ============================================================================
 set -e
 # Colors for output
 RED='\033[0;31m'
 GREEN='\033[0;32m'
 YELLOW='\033[1;33m'
 BLUE='\033[0;34m'
 NC='\033[0m' # No Color
 # Configuration
 DB_HOST="${POSTGRES_HOST:-localhost}"
 DB_PORT="${POSTGRES_PORT:-5432}"
 DB_NAME="${POSTGRES_DB:-fraud_detection}"
 DB_USER="${POSTGRES_USER:-fraud_analyst}"
 DB_PASSWORD="${POSTGRES_PASSWORD:-SecurePass123!}"
 # Data volumes
 NUM_CUSTOMERS=100000
 NUM_ACCOUNTS=150000
 NUM_MERCHANTS=50000
 NUM_DEVICES=75000
 NUM_CARDS=200000
 NUM_TRANSACTIONS=5000000
 FRAUD_PERCENTAGE=7  # 7% of transactions will be fraudulent
 echo -e "${BLUE}============================================================================${NC}"
 echo -e "${BLUE}Financial Fraud Detection Database - Data Generation${NC}"
 echo -e "${BLUE}============================================================================${NC}"
 echo -e "Target Database: ${GREEN}$DB_NAME@$DB_HOST:$DB_PORT${NC}"
 echo -e "Customers: ${GREEN}$NUM_CUSTOMERS${NC}"
 echo -e "Accounts: ${GREEN}$NUM_ACCOUNTS${NC}"
 echo -e "Merchants: ${GREEN}$NUM_MERCHANTS${NC}"
 echo -e "Cards: ${GREEN}$NUM_CARDS${NC}"
 echo -e "Transactions: ${GREEN}$NUM_TRANSACTIONS${NC} (${YELLOW}${FRAUD_PERCENTAGE}% fraudulent${NC})"
 echo -e "${BLUE}============================================================================${NC}"
 echo ""
 echo -e "${RED}WARNING: This will DELETE all existing data and regenerate it!${NC}"
 echo -e "${YELLOW}Press Ctrl+C within 5 seconds to cancel...${NC}"
 sleep 5
 echo ""
 # Function to execute SQL
 execute_sql() {
    PGPASSWORD=$DB_PASSWORD psql -h $DB_HOST -p $DB_PORT -U $DB_USER -d $DB_NAME -c "$1" 2>&1
 }
 # Function to execute SQL file
 execute_sql_file() {
    PGPASSWORD=$DB_PASSWORD psql -h $DB_HOST -p $DB_PORT -U $DB_USER -d $DB_NAME -f "$1" 2>&1
 }
 # Get script directory
 SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
 # Clear existing data (preserve reference tables)
 echo -e "${YELLOW}[0/9] Clearing existing data...${NC}"
 execute_sql "TRUNCATE TABLE audit_log CASCADE;"
 execute_sql "TRUNCATE TABLE suspicious_activity_reports CASCADE;"
 execute_sql "TRUNCATE TABLE case_alerts CASCADE;"
 execute_sql "TRUNCATE TABLE case_transactions CASCADE;"
 execute_sql "TRUNCATE TABLE fraud_cases CASCADE;"
 execute_sql "TRUNCATE TABLE alerts CASCADE;"
 execute_sql "TRUNCATE TABLE transfers CASCADE;"
 execute_sql "TRUNCATE TABLE beneficiaries CASCADE;"
 execute_sql "TRUNCATE TABLE transactions CASCADE;"
 execute_sql "TRUNCATE TABLE login_sessions CASCADE;"
 execute_sql "TRUNCATE TABLE devices RESTART IDENTITY CASCADE;"
 execute_sql "TRUNCATE TABLE cards RESTART IDENTITY CASCADE;"
 execute_sql "TRUNCATE TABLE accounts RESTART IDENTITY CASCADE;"
 execute_sql "TRUNCATE TABLE customer_relationships RESTART IDENTITY CASCADE;"
 execute_sql "TRUNCATE TABLE customers RESTART IDENTITY CASCADE;"
 execute_sql "TRUNCATE TABLE merchants RESTART IDENTITY CASCADE;"
 echo -e "${GREEN}✓ Existing data cleared${NC}"
 echo ""
 # Load geographic reference data
 echo -e "${YELLOW}[1/9] Loading geographic reference data...${NC}"
 execute_sql_file "$SCRIPT_DIR/reference/load_geographic_data.sql" > /dev/null
 echo -e "${GREEN}✓ Geographic data loaded (100 US cities, 210 world cities)${NC}"
 echo ""
 echo -e "${YELLOW}[2/9] Generating Customers...${NC}"
 cat > /tmp/generate_customers.sql << 'EOF'
 -- Generate customers with realistic geographic data
 INSERT INTO customers (
    first_name, last_name, email, phone, date_of_birth, ssn_hash,
    address_line1, city, state, postal_code, country_id,
    registration_date, kyc_status, risk_score, is_pep, is_active
 )
 SELECT
    'Customer' || gs.id AS first_name,
    'User' || gs.id AS last_name,
    'customer' || gs.id || '@email.com' AS email,
    '+1' || LPAD((1000000000 + (gs.id % 9000000000))::TEXT, 10, '0') AS phone,
    DATE '1950-01-01' + (random() * 25000)::INT AS date_of_birth,
    encode(digest('SSN' || gs.id::TEXT, 'sha256'), 'hex') AS ssn_hash,
    (gs.id % 10000) || ' Main Street' AS address_line1,
    -- Use real US cities from temp table
    (SELECT city FROM temp_us_cities WHERE id = ((gs.id % 100) + 1)) AS city,
    (SELECT state_code FROM temp_us_cities WHERE id = ((gs.id % 100) + 1)) AS state,
    LPAD((10000 + (gs.id % 90000))::TEXT, 5, '0') AS postal_code,
    CASE
        WHEN random() < 0.85 THEN 1  -- 85% US
        WHEN random() < 0.90 THEN 2  -- 5% Canada
        WHEN random() < 0.95 THEN 3  -- 5% UK
        ELSE (3 + (gs.id % 37))      -- 5% other countries
    END AS country_id,
    TIMESTAMP '2020-01-01' + (random() * 1460)::INT * INTERVAL '1 day' AS registration_date,
    CASE
        WHEN random() < 0.90 THEN 'VERIFIED'
        WHEN random() < 0.95 THEN 'PENDING'
        ELSE 'REJECTED'
    END AS kyc_status,
    (random() * 100)::DECIMAL(5,2) AS risk_score,
    random() < 0.02 AS is_pep,  -- 2% are PEPs
    random() < 0.98 AS is_active  -- 98% active
 FROM generate_series(1, 100000) AS gs(id);
 EOF
 execute_sql_file /tmp/generate_customers.sql > /dev/null
 echo -e "${GREEN}✓ Generated $NUM_CUSTOMERS customers${NC}"
 echo ""
 echo -e "${YELLOW}[3/9] Generating Accounts...${NC}"
 cat > /tmp/generate_accounts.sql << 'EOF'
 -- Generate accounts (1-2 accounts per customer on average)
 INSERT INTO accounts (
    customer_id, account_number, account_type, currency,
    opening_date, status, current_balance, available_balance,
    credit_limit, overdraft_limit, is_primary
 )
 SELECT
    (gs.id % 100000) + 1 AS customer_id,
    'ACC' || LPAD(gs.id::TEXT, 12, '0') AS account_number,
    CASE (gs.id % 10)
        WHEN 0 THEN 'CHECKING'
        WHEN 1 THEN 'CHECKING'
        WHEN 2 THEN 'CHECKING'
        WHEN 3 THEN 'SAVINGS'
        WHEN 4 THEN 'SAVINGS'
        WHEN 5 THEN 'CREDIT'
        WHEN 6 THEN 'CREDIT'
        WHEN 7 THEN 'INVESTMENT'
        ELSE 'CHECKING'
    END AS account_type,
    'USD' AS currency,
    DATE '2020-01-01' + (random() * 1460)::INT AS opening_date,
    CASE 
        WHEN random() < 0.95 THEN 'ACTIVE'
        WHEN random() < 0.98 THEN 'SUSPENDED'
        ELSE 'CLOSED'
    END AS status,
    (random() * 50000)::DECIMAL(15,2) AS current_balance,
    (random() * 50000)::DECIMAL(15,2) AS available_balance,
    CASE 
        WHEN (gs.id % 10) IN (5, 6) THEN (5000 + random() * 45000)::DECIMAL(15,2)
        ELSE NULL
    END AS credit_limit,
    CASE 
        WHEN (gs.id % 10) IN (0, 1, 2) THEN (random() * 1000)::DECIMAL(15,2)
        ELSE 0
    END AS overdraft_limit,
    (gs.id % 2) = 0 AS is_primary
 FROM generate_series(1, 150000) AS gs(id);
 EOF
 execute_sql_file /tmp/generate_accounts.sql > /dev/null
 echo -e "${GREEN}✓ Generated $NUM_ACCOUNTS accounts${NC}"
 echo ""
 echo -e "${YELLOW}[4/9] Generating Merchants...${NC}"
 cat > /tmp/generate_merchants.sql << 'EOF'
 -- Generate merchants with realistic geographic data
 INSERT INTO merchants (
    merchant_name, merchant_code, category_id, country_id,
    city, registration_date, status, risk_rating, is_verified
 )
 SELECT
    CASE (gs.id % 15)
        WHEN 0 THEN 'Walmart Store #' || gs.id
        WHEN 1 THEN 'Amazon Marketplace #' || gs.id
        WHEN 2 THEN 'Shell Gas Station #' || gs.id
        WHEN 3 THEN 'McDonalds #' || gs.id
        WHEN 4 THEN 'Starbucks #' || gs.id
        WHEN 5 THEN 'Target Store #' || gs.id
        WHEN 6 THEN 'Best Buy #' || gs.id
        WHEN 7 THEN 'CVS Pharmacy #' || gs.id
        WHEN 8 THEN 'Home Depot #' || gs.id
        WHEN 9 THEN 'Costco #' || gs.id
        WHEN 10 THEN 'Apple Store #' || gs.id
        WHEN 11 THEN 'Marriott Hotel #' || gs.id
        WHEN 12 THEN 'Delta Airlines #' || gs.id
        WHEN 13 THEN 'Online Casino #' || gs.id
        ELSE 'Merchant #' || gs.id
    END AS merchant_name,
    'MER' || LPAD(gs.id::TEXT, 10, '0') AS merchant_code,
    ((gs.id % 35) + 1) AS category_id,
    CASE
        WHEN random() < 0.80 THEN 1  -- 80% US merchants
        WHEN random() < 0.90 THEN 2  -- 10% Canada
        ELSE (3 + (gs.id % 37))      -- 10% international
    END AS country_id,
    -- Use real cities based on country
    CASE
        WHEN random() < 0.80 THEN (SELECT city FROM temp_us_cities WHERE id = ((gs.id % 100) + 1))
        ELSE (SELECT city FROM temp_world_cities WHERE id = ((gs.id % 210) + 1))
    END AS city,
    DATE '2015-01-01' + (random() * 3000)::INT AS registration_date,
    CASE
        WHEN random() < 0.95 THEN 'ACTIVE'
        WHEN random() < 0.98 THEN 'SUSPENDED'
        ELSE 'BLACKLISTED'
    END AS status,
    CASE
        WHEN (gs.id % 35) + 1 IN (21, 22, 23, 24, 25, 26, 31, 32, 33, 34, 35) THEN
            CASE
                WHEN random() < 0.5 THEN 'HIGH'
                ELSE 'CRITICAL'
            END
        WHEN random() < 0.80 THEN 'LOW'
        ELSE 'MEDIUM'
    END AS risk_rating,
    random() < 0.90 AS is_verified
 FROM generate_series(1, 50000) AS gs(id);
 EOF
 execute_sql_file /tmp/generate_merchants.sql > /dev/null
 echo -e "${GREEN}✓ Generated $NUM_MERCHANTS merchants${NC}"
 echo ""
 echo -e "${YELLOW}[5/9] Generating Devices...${NC}"
 cat > /tmp/generate_devices.sql << 'EOF'
 -- Generate devices
 INSERT INTO devices (
    device_fingerprint, device_type, os_name, os_version,
    browser_name, browser_version, is_trusted, is_blacklisted
 )
 SELECT
    encode(digest('DEVICE' || gs.id::TEXT, 'sha256'), 'hex') AS device_fingerprint,
    CASE (gs.id % 4)
        WHEN 0 THEN 'MOBILE'
        WHEN 1 THEN 'DESKTOP'
        WHEN 2 THEN 'TABLET'
        ELSE 'MOBILE'
    END AS device_type,
    CASE (gs.id % 5)
        WHEN 0 THEN 'iOS'
        WHEN 1 THEN 'Android'
        WHEN 2 THEN 'Windows'
        WHEN 3 THEN 'macOS'
        ELSE 'Linux'
    END AS os_name,
    CASE (gs.id % 5)
        WHEN 0 THEN '15.0'
        WHEN 1 THEN '12.0'
        WHEN 2 THEN '11.0'
        WHEN 3 THEN '13.0'
        ELSE '10.0'
    END AS os_version,
    CASE (gs.id % 4)
        WHEN 0 THEN 'Chrome'
        WHEN 1 THEN 'Safari'
        WHEN 2 THEN 'Firefox'
        ELSE 'Edge'
    END AS browser_name,
    '100.0' AS browser_version,
    random() < 0.85 AS is_trusted,
    random() < 0.03 AS is_blacklisted
 FROM generate_series(1, 75000) AS gs(id);
 EOF
 execute_sql_file /tmp/generate_devices.sql > /dev/null
 echo -e "${GREEN}✓ Generated $NUM_DEVICES devices${NC}"
 echo ""
 echo -e "${YELLOW}[6/9] Generating Cards...${NC}"
 cat > /tmp/generate_cards.sql << 'EOF'
 -- Generate cards (1-2 cards per account on average)
 INSERT INTO cards (
    account_id, card_number_hash, card_last_four, card_type, card_network,
    issue_date, expiry_date, cvv_hash, status, daily_limit, monthly_limit,
    is_contactless, is_international
 )
 SELECT
    ((gs.id - 1) % 150000) + 1 AS account_id,
    encode(digest('CARD' || gs.id::TEXT, 'sha256'), 'hex') AS card_number_hash,
    LPAD((gs.id % 10000)::TEXT, 4, '0') AS card_last_four,
    CASE (gs.id % 4)
        WHEN 0 THEN 'DEBIT'
        WHEN 1 THEN 'CREDIT'
        WHEN 2 THEN 'DEBIT'
        ELSE 'CREDIT'
    END AS card_type,
    CASE (gs.id % 4)
        WHEN 0 THEN 'VISA'
        WHEN 1 THEN 'MASTERCARD'
        WHEN 2 THEN 'AMEX'
        ELSE 'DISCOVER'
    END AS card_network,
    DATE '2020-01-01' + (random() * 1000)::INT AS issue_date,
    DATE '2025-01-01' + (random() * 1825)::INT AS expiry_date,
    encode(digest('CVV' || gs.id::TEXT, 'sha256'), 'hex') AS cvv_hash,
    CASE
        WHEN random() < 0.95 THEN 'ACTIVE'
        WHEN random() < 0.97 THEN 'BLOCKED'
        WHEN random() < 0.99 THEN 'LOST'
        ELSE 'STOLEN'
    END AS status,
    (1000 + random() * 9000)::DECIMAL(10,2) AS daily_limit,
    (10000 + random() * 90000)::DECIMAL(12,2) AS monthly_limit,
    random() < 0.90 AS is_contactless,
    random() < 0.30 AS is_international
 FROM generate_series(1, 200000) AS gs(id);
 EOF
 execute_sql_file /tmp/generate_cards.sql > /dev/null
 echo -e "${GREEN}✓ Generated $NUM_CARDS cards${NC}"
 echo ""
 echo -e "${YELLOW}[7/9] Generating Login Sessions...${NC}"
 cat > /tmp/generate_sessions.sql << 'EOF'
 -- Generate login sessions with realistic geographic data
 INSERT INTO login_sessions (
    customer_id, device_id, ip_address, country_id, city,
    login_timestamp, logout_timestamp, session_duration_seconds,
    is_successful, risk_score
 )
 SELECT
    ((gs.id - 1) % 100000) + 1 AS customer_id,
    ((gs.id - 1) % 75000) + 1 AS device_id,
    ('192.168.' || ((gs.id % 255) + 1) || '.' || ((gs.id % 255) + 1))::INET AS ip_address,
    CASE
        WHEN random() < 0.85 THEN 1
        ELSE ((gs.id % 40) + 1)
    END AS country_id,
    -- Use real cities
    CASE
        WHEN random() < 0.85 THEN (SELECT city FROM temp_us_cities WHERE id = ((gs.id % 100) + 1))
        ELSE (SELECT city FROM temp_world_cities WHERE id = ((gs.id % 210) + 1))
    END AS city,
    TIMESTAMP '2023-01-01' + (random() * 730)::INT * INTERVAL '1 day' + (random() * 86400)::INT * INTERVAL '1 second' AS login_timestamp,
    TIMESTAMP '2023-01-01' + (random() * 730)::INT * INTERVAL '1 day' + (random() * 86400)::INT * INTERVAL '1 second' + (random() * 7200)::INT * INTERVAL '1 second' AS logout_timestamp,
    (300 + random() * 7200)::INT AS session_duration_seconds,
    random() < 0.98 AS is_successful,
    (random() * 100)::DECIMAL(5,2) AS risk_score
 FROM generate_series(1, 500000) AS gs(id);
 EOF
 execute_sql_file /tmp/generate_sessions.sql > /dev/null
 echo -e "${GREEN}✓ Generated 500,000 login sessions${NC}"
 echo ""
 echo -e "${YELLOW}[8/9] Generating Transactions (this may take a while)...${NC}"
 echo -e "${BLUE}This step generates $NUM_TRANSACTIONS transactions with fraud patterns${NC}"
 # Generate transactions in batches to avoid memory issues
 BATCH_SIZE=500000
 NUM_BATCHES=$((NUM_TRANSACTIONS / BATCH_SIZE))
 for batch in $(seq 1 $NUM_BATCHES); do
    START_ID=$(( (batch - 1) * BATCH_SIZE + 1 ))
    END_ID=$(( batch * BATCH_SIZE ))
    echo -e "${BLUE}  Batch $batch/$NUM_BATCHES (transactions $START_ID to $END_ID)...${NC}"
    cat > /tmp/generate_transactions_batch.sql << EOF
 -- Generate transactions batch
 INSERT INTO transactions (
    account_id, type_id, transaction_date, amount, currency,
    merchant_id, card_id, device_id, ip_address, country_id, city,
    description, reference_number, status, is_online, is_international,
    is_card_present, fraud_score, is_flagged
 )
 SELECT
    ((gs.id - 1) % 150000) + 1 AS account_id,
    ((gs.id % 15) + 1) AS type_id,
    TIMESTAMP '2023-01-01' + (random() * 730)::INT * INTERVAL '1 day' + (random() * 86400)::INT * INTERVAL '1 second' AS transaction_date,
    CASE
        WHEN random() < 0.60 THEN (5 + random() * 95)::DECIMAL(15,2)
        WHEN random() < 0.85 THEN (100 + random() * 400)::DECIMAL(15,2)
        WHEN random() < 0.95 THEN (500 + random() * 2000)::DECIMAL(15,2)
        WHEN random() < 0.98 THEN (2500 + random() * 7500)::DECIMAL(15,2)
        ELSE (10000 + random() * 90000)::DECIMAL(15,2)
    END AS amount,
    'USD' AS currency,
    CASE
        WHEN ((gs.id % 15) + 1) IN (1, 7, 13) THEN ((gs.id % 50000) + 1)
        ELSE NULL
    END AS merchant_id,
    CASE
        WHEN ((gs.id % 15) + 1) IN (1, 7, 13) THEN ((gs.id % 200000) + 1)
        ELSE NULL
    END AS card_id,
    ((gs.id % 75000) + 1) AS device_id,
    ('10.' || ((gs.id % 255) + 1) || '.' || ((gs.id % 255) + 1) || '.' || ((gs.id % 255) + 1))::INET AS ip_address,
    CASE
        WHEN random() < 0.90 THEN 1
        ELSE ((gs.id % 40) + 1)
    END AS country_id,
    -- Use real cities based on country
    CASE
        WHEN random() < 0.90 THEN (SELECT city FROM temp_us_cities WHERE id = ((gs.id % 100) + 1))
        ELSE (SELECT city FROM temp_world_cities WHERE id = ((gs.id % 210) + 1))
    END AS city,
    'Transaction #' || gs.id AS description,
    'REF' || LPAD(gs.id::TEXT, 15, '0') AS reference_number,
    CASE
        WHEN random() < 0.95 THEN 'COMPLETED'
        WHEN random() < 0.98 THEN 'PENDING'
        ELSE 'FAILED'
    END AS status,
    random() < 0.70 AS is_online,
    random() < 0.10 AS is_international,
    random() < 0.30 AS is_card_present,
    (random() * 100)::DECIMAL(5,2) AS fraud_score,
    random() < 0.07 AS is_flagged
 FROM generate_series($START_ID, $END_ID) AS gs(id);
 EOF
    execute_sql_file /tmp/generate_transactions_batch.sql > /dev/null
    echo -e "${GREEN}  ✓ Batch $batch/$NUM_BATCHES completed${NC}"
 done
 echo -e "${GREEN}✓ Generated $NUM_TRANSACTIONS transactions${NC}"
 echo ""
 echo -e "${YELLOW}[9/9] Generating Fraud Cases and Alerts...${NC}"
 # Generate alerts for flagged transactions
 execute_sql "
 INSERT INTO alerts (transaction_id, customer_id, account_id, alert_type, severity, description, risk_score, status)
 SELECT
    t.transaction_id,
    a.customer_id,
    t.account_id,
    CASE
        WHEN t.amount > 10000 THEN 'AMOUNT_ANOMALY'
        WHEN t.is_international THEN 'GEOGRAPHIC_ANOMALY'
        WHEN t.fraud_score > 80 THEN 'MERCHANT_RISK'
        ELSE 'VELOCITY_CHECK'
    END AS alert_type,
    CASE
        WHEN t.fraud_score > 80 THEN 'CRITICAL'
        WHEN t.fraud_score > 60 THEN 'HIGH'
        ELSE 'MEDIUM'
    END AS severity,
    'Suspicious transaction detected: ' || t.description AS description,
    t.fraud_score,
    CASE
        WHEN random() < 0.30 THEN 'CLOSED'
        WHEN random() < 0.50 THEN 'FALSE_POSITIVE'
        WHEN random() < 0.70 THEN 'INVESTIGATING'
        ELSE 'OPEN'
    END AS status
 FROM transactions t
 JOIN accounts a ON t.account_id = a.account_id
 WHERE t.is_flagged = TRUE
 LIMIT 50000;
 " > /dev/null
 echo -e "${GREEN}✓ Generated alerts for flagged transactions${NC}"
 # Generate fraud cases
 execute_sql "
 INSERT INTO fraud_cases (
    case_number, customer_id, account_id, fraud_type_id,
    detection_date, detection_method, amount_lost, status, priority
 )
 SELECT
    'CASE' || LPAD(ROW_NUMBER() OVER (ORDER BY a.alert_id)::TEXT, 10, '0') AS case_number,
    a.customer_id,
    a.account_id,
    ((a.alert_id % 20) + 1) AS fraud_type_id,
    a.alert_date AS detection_date,
    CASE
        WHEN random() < 0.70 THEN 'AUTOMATED'
        WHEN random() < 0.85 THEN 'MANUAL_REVIEW'
        ELSE 'CUSTOMER_REPORT'
    END AS detection_method,
    (random() * 50000)::DECIMAL(15,2) AS amount_lost,
    CASE
        WHEN random() < 0.40 THEN 'RESOLVED'
        WHEN random() < 0.60 THEN 'INVESTIGATING'
        ELSE 'OPEN'
    END AS status,
    CASE
        WHEN a.severity = 'CRITICAL' THEN 'CRITICAL'
        WHEN a.severity = 'HIGH' THEN 'HIGH'
        ELSE 'MEDIUM'
    END AS priority
 FROM alerts a
 WHERE a.status = 'CONFIRMED_FRAUD'
   OR (a.severity IN ('CRITICAL', 'HIGH') AND random() < 0.20)
 LIMIT 5000;
 " > /dev/null
 echo -e "${GREEN}✓ Generated fraud cases${NC}"
 echo ""
 # Final statistics
 echo -e "${BLUE}============================================================================${NC}"
 echo -e "${GREEN}✓ Data generation completed successfully!${NC}"
 echo -e "${BLUE}============================================================================${NC}"
 echo ""
 echo -e "${YELLOW}Database Statistics:${NC}"
 customer_count=$(execute_sql "SELECT COUNT(*) FROM customers;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Customers: ${GREEN}$customer_count${NC}"
 account_count=$(execute_sql "SELECT COUNT(*) FROM accounts;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Accounts: ${GREEN}$account_count${NC}"
 merchant_count=$(execute_sql "SELECT COUNT(*) FROM merchants;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Merchants: ${GREEN}$merchant_count${NC}"
 card_count=$(execute_sql "SELECT COUNT(*) FROM cards;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Cards: ${GREEN}$card_count${NC}"
 transaction_count=$(execute_sql "SELECT COUNT(*) FROM transactions;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Transactions: ${GREEN}$transaction_count${NC}"
 alert_count=$(execute_sql "SELECT COUNT(*) FROM alerts;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Alerts: ${GREEN}$alert_count${NC}"
 case_count=$(execute_sql "SELECT COUNT(*) FROM fraud_cases;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Fraud Cases: ${GREEN}$case_count${NC}"
 echo ""
 echo -e "${YELLOW}Ready for SQL learning and fraud investigation!${NC}"
 echo -e "Access DB-UI at: ${BLUE}http://localhost:3000${NC}"
 echo ""
@@ -0,0 +1,329 @@
 -- ============================================================================
 -- Load Geographic Reference Data into Temporary Tables
 -- ============================================================================
 -- This script creates temporary tables with realistic city/state/country data
 -- to be used during data generation
 -- ============================================================================
 -- Drop temporary tables if they exist
 DROP TABLE IF EXISTS temp_us_cities CASCADE;
 DROP TABLE IF EXISTS temp_world_cities CASCADE;
 -- Create temporary table for US cities
 CREATE TEMPORARY TABLE temp_us_cities (
    id SERIAL PRIMARY KEY,
    city VARCHAR(100) NOT NULL,
    state VARCHAR(100) NOT NULL,
    state_code CHAR(2) NOT NULL
 );
 -- Create temporary table for world cities
 CREATE TEMPORARY TABLE temp_world_cities (
    id SERIAL PRIMARY KEY,
    city VARCHAR(100) NOT NULL,
    country VARCHAR(100) NOT NULL,
    country_code CHAR(2) NOT NULL
 );
 -- Load US cities data
 COPY temp_us_cities(city, state, state_code) FROM STDIN WITH (FORMAT CSV, HEADER true);
 New York,New York,NY
 Los Angeles,California,CA
 Chicago,Illinois,IL
 Houston,Texas,TX
 Phoenix,Arizona,AZ
 Philadelphia,Pennsylvania,PA
 San Antonio,Texas,TX
 San Diego,California,CA
 Dallas,Texas,TX
 San Jose,California,CA
 Austin,Texas,TX
 Jacksonville,Florida,FL
 Fort Worth,Texas,TX
 Columbus,Ohio,OH
 Charlotte,North Carolina,NC
 San Francisco,California,CA
 Indianapolis,Indiana,IN
 Seattle,Washington,WA
 Denver,Colorado,CO
 Boston,Massachusetts,MA
 Nashville,Tennessee,TN
 Detroit,Michigan,MI
 Portland,Oregon,OR
 Las Vegas,Nevada,NV
 Memphis,Tennessee,TN
 Louisville,Kentucky,KY
 Baltimore,Maryland,MD
 Milwaukee,Wisconsin,WI
 Albuquerque,New Mexico,NM
 Tucson,Arizona,AZ
 Fresno,California,CA
 Sacramento,California,CA
 Kansas City,Missouri,MO
 Mesa,Arizona,AZ
 Atlanta,Georgia,GA
 Omaha,Nebraska,NE
 Colorado Springs,Colorado,CO
 Raleigh,North Carolina,NC
 Miami,Florida,FL
 Long Beach,California,CA
 Virginia Beach,Virginia,VA
 Oakland,California,CA
 Minneapolis,Minnesota,MN
 Tampa,Florida,FL
 Tulsa,Oklahoma,OK
 Arlington,Texas,TX
 New Orleans,Louisiana,LA
 Wichita,Kansas,KS
 Cleveland,Ohio,OH
 Bakersfield,California,CA
 Aurora,Colorado,CO
 Anaheim,California,CA
 Honolulu,Hawaii,HI
 Santa Ana,California,CA
 Riverside,California,CA
 Corpus Christi,Texas,TX
 Lexington,Kentucky,KY
 Stockton,California,CA
 Henderson,Nevada,NV
 Saint Paul,Minnesota,MN
 St. Louis,Missouri,MO
 Cincinnati,Ohio,OH
 Pittsburgh,Pennsylvania,PA
 Greensboro,North Carolina,NC
 Anchorage,Alaska,AK
 Plano,Texas,TX
 Lincoln,Nebraska,NE
 Orlando,Florida,FL
 Irvine,California,CA
 Newark,New Jersey,NJ
 Durham,North Carolina,NC
 Chula Vista,California,CA
 Toledo,Ohio,OH
 Fort Wayne,Indiana,IN
 St. Petersburg,Florida,FL
 Laredo,Texas,TX
 Jersey City,New Jersey,NJ
 Chandler,Arizona,AZ
 Madison,Wisconsin,WI
 Lubbock,Texas,TX
 Scottsdale,Arizona,AZ
 Reno,Nevada,NV
 Buffalo,New York,NY
 Gilbert,Arizona,AZ
 Glendale,Arizona,AZ
 North Las Vegas,Nevada,NV
 Winston-Salem,North Carolina,NC
 Chesapeake,Virginia,VA
 Norfolk,Virginia,VA
 Fremont,California,CA
 Garland,Texas,TX
 Irving,Texas,TX
 Hialeah,Florida,FL
 Richmond,Virginia,VA
 Boise,Idaho,ID
 Spokane,Washington,WA
 Baton Rouge,Louisiana,LA
 \.
 -- Load world cities data
 COPY temp_world_cities(city, country, country_code) FROM STDIN WITH (FORMAT CSV, HEADER true);
 Toronto,Canada,CA
 Vancouver,Canada,CA
 Montreal,Canada,CA
 Calgary,Canada,CA
 Ottawa,Canada,CA
 London,United Kingdom,GB
 Manchester,United Kingdom,GB
 Birmingham,United Kingdom,GB
 Edinburgh,United Kingdom,GB
 Glasgow,United Kingdom,GB
 Berlin,Germany,DE
 Munich,Germany,DE
 Hamburg,Germany,DE
 Frankfurt,Germany,DE
 Cologne,Germany,DE
 Paris,France,FR
 Lyon,France,FR
 Marseille,France,FR
 Toulouse,France,FR
 Nice,France,FR
 Rome,Italy,IT
 Milan,Italy,IT
 Naples,Italy,IT
 Turin,Italy,IT
 Florence,Italy,IT
 Madrid,Spain,ES
 Barcelona,Spain,ES
 Valencia,Spain,ES
 Seville,Spain,ES
 Bilbao,Spain,ES
 Sydney,Australia,AU
 Melbourne,Australia,AU
 Brisbane,Australia,AU
 Perth,Australia,AU
 Adelaide,Australia,AU
 Tokyo,Japan,JP
 Osaka,Japan,JP
 Kyoto,Japan,JP
 Yokohama,Japan,JP
 Nagoya,Japan,JP
 Beijing,China,CN
 Shanghai,China,CN
 Guangzhou,China,CN
 Shenzhen,China,CN
 Chengdu,China,CN
 Mumbai,India,IN
 Delhi,India,IN
 Bangalore,India,IN
 Hyderabad,India,IN
 Chennai,India,IN
 Sao Paulo,Brazil,BR
 Rio de Janeiro,Brazil,BR
 Brasilia,Brazil,BR
 Salvador,Brazil,BR
 Fortaleza,Brazil,BR
 Mexico City,Mexico,MX
 Guadalajara,Mexico,MX
 Monterrey,Mexico,MX
 Puebla,Mexico,MX
 Tijuana,Mexico,MX
 Moscow,Russia,RU
 Saint Petersburg,Russia,RU
 Novosibirsk,Russia,RU
 Yekaterinburg,Russia,RU
 Kazan,Russia,RU
 Lagos,Nigeria,NG
 Kano,Nigeria,NG
 Ibadan,Nigeria,NG
 Abuja,Nigeria,NG
 Port Harcourt,Nigeria,NG
 Karachi,Pakistan,PK
 Lahore,Pakistan,PK
 Islamabad,Pakistan,PK
 Rawalpindi,Pakistan,PK
 Faisalabad,Pakistan,PK
 Tehran,Iran,IR
 Mashhad,Iran,IR
 Isfahan,Iran,IR
 Karaj,Iran,IR
 Tabriz,Iran,IR
 Pyongyang,North Korea,KP
 Hamhung,North Korea,KP
 Chongjin,North Korea,KP
 Nampo,North Korea,KP
 Wonsan,North Korea,KP
 Damascus,Syria,SY
 Aleppo,Syria,SY
 Homs,Syria,SY
 Latakia,Syria,SY
 Hama,Syria,SY
 Caracas,Venezuela,VE
 Maracaibo,Venezuela,VE
 Valencia,Venezuela,VE
 Barquisimeto,Venezuela,VE
 Maracay,Venezuela,VE
 Havana,Cuba,CU
 Santiago de Cuba,Cuba,CU
 Camaguey,Cuba,CU
 Holguin,Cuba,CU
 Santa Clara,Cuba,CU
 Yangon,Myanmar,MM
 Mandalay,Myanmar,MM
 Naypyidaw,Myanmar,MM
 Mawlamyine,Myanmar,MM
 Bago,Myanmar,MM
 Kabul,Afghanistan,AF
 Kandahar,Afghanistan,AF
 Herat,Afghanistan,AF
 Mazar-i-Sharif,Afghanistan,AF
 Jalalabad,Afghanistan,AF
 Baghdad,Iraq,IQ
 Basra,Iraq,IQ
 Mosul,Iraq,IQ
 Erbil,Iraq,IQ
 Kirkuk,Iraq,IQ
 Tripoli,Libya,LY
 Benghazi,Libya,LY
 Misrata,Libya,LY
 Zawiya,Libya,LY
 Bayda,Libya,LY
 Khartoum,Sudan,SD
 Omdurman,Sudan,SD
 Port Sudan,Sudan,SD
 Kassala,Sudan,SD
 Nyala,Sudan,SD
 Mogadishu,Somalia,SO
 Hargeisa,Somalia,SO
 Bosaso,Somalia,SO
 Kismayo,Somalia,SO
 Merca,Somalia,SO
 Sanaa,Yemen,YE
 Aden,Yemen,YE
 Taiz,Yemen,YE
 Hodeidah,Yemen,YE
 Ibb,Yemen,YE
 Harare,Zimbabwe,ZW
 Bulawayo,Zimbabwe,ZW
 Chitungwiza,Zimbabwe,ZW
 Mutare,Zimbabwe,ZW
 Gweru,Zimbabwe,ZW
 Amsterdam,Netherlands,NL
 Rotterdam,Netherlands,NL
 The Hague,Netherlands,NL
 Utrecht,Netherlands,NL
 Eindhoven,Netherlands,NL
 Stockholm,Sweden,SE
 Gothenburg,Sweden,SE
 Malmo,Sweden,SE
 Uppsala,Sweden,SE
 Vasteras,Sweden,SE
 Oslo,Norway,NO
 Bergen,Norway,NO
 Trondheim,Norway,NO
 Stavanger,Norway,NO
 Drammen,Norway,NO
 Copenhagen,Denmark,DK
 Aarhus,Denmark,DK
 Odense,Denmark,DK
 Aalborg,Denmark,DK
 Esbjerg,Denmark,DK
 Helsinki,Finland,FI
 Espoo,Finland,FI
 Tampere,Finland,FI
 Vantaa,Finland,FI
 Oulu,Finland,FI
 Zurich,Switzerland,CH
 Geneva,Switzerland,CH
 Basel,Switzerland,CH
 Lausanne,Switzerland,CH
 Bern,Switzerland,CH
 Singapore,Singapore,SG
 Hong Kong,Hong Kong,HK
 Kowloon,Hong Kong,HK
 Seoul,South Korea,KR
 Busan,South Korea,KR
 Incheon,South Korea,KR
 Daegu,South Korea,KR
 Daejeon,South Korea,KR
 Taipei,Taiwan,TW
 Kaohsiung,Taiwan,TW
 Taichung,Taiwan,TW
 Tainan,Taiwan,TW
 Hsinchu,Taiwan,TW
 Auckland,New Zealand,NZ
 Wellington,New Zealand,NZ
 Christchurch,New Zealand,NZ
 Hamilton,New Zealand,NZ
 Tauranga,New Zealand,NZ
 \.
 -- Create indexes for faster lookups
 CREATE INDEX idx_temp_us_cities_id ON temp_us_cities(id);
 CREATE INDEX idx_temp_world_cities_id ON temp_world_cities(id);
 CREATE INDEX idx_temp_world_cities_country_code ON temp_world_cities(country_code);
 -- Show counts
 SELECT 'US Cities loaded: ' || COUNT(*) FROM temp_us_cities;
 SELECT 'World Cities loaded: ' || COUNT(*) FROM temp_world_cities;
@@ -0,0 +1,99 @@
 city,state,state_code
 New York,New York,NY
 Los Angeles,California,CA
 Chicago,Illinois,IL
 Houston,Texas,TX
 Phoenix,Arizona,AZ
 Philadelphia,Pennsylvania,PA
 San Antonio,Texas,TX
 San Diego,California,CA
 Dallas,Texas,TX
 San Jose,California,CA
 Austin,Texas,TX
 Jacksonville,Florida,FL
 Fort Worth,Texas,TX
 Columbus,Ohio,OH
 Charlotte,North Carolina,NC
 San Francisco,California,CA
 Indianapolis,Indiana,IN
 Seattle,Washington,WA
 Denver,Colorado,CO
 Boston,Massachusetts,MA
 Nashville,Tennessee,TN
 Detroit,Michigan,MI
 Portland,Oregon,OR
 Las Vegas,Nevada,NV
 Memphis,Tennessee,TN
 Louisville,Kentucky,KY
 Baltimore,Maryland,MD
 Milwaukee,Wisconsin,WI
 Albuquerque,New Mexico,NM
 Tucson,Arizona,AZ
 Fresno,California,CA
 Sacramento,California,CA
 Kansas City,Missouri,MO
 Mesa,Arizona,AZ
 Atlanta,Georgia,GA
 Omaha,Nebraska,NE
 Colorado Springs,Colorado,CO
 Raleigh,North Carolina,NC
 Miami,Florida,FL
 Long Beach,California,CA
 Virginia Beach,Virginia,VA
 Oakland,California,CA
 Minneapolis,Minnesota,MN
 Tampa,Florida,FL
 Tulsa,Oklahoma,OK
 Arlington,Texas,TX
 New Orleans,Louisiana,LA
 Wichita,Kansas,KS
 Cleveland,Ohio,OH
 Bakersfield,California,CA
 Aurora,Colorado,CO
 Anaheim,California,CA
 Honolulu,Hawaii,HI
 Santa Ana,California,CA
 Riverside,California,CA
 Corpus Christi,Texas,TX
 Lexington,Kentucky,KY
 Stockton,California,CA
 Henderson,Nevada,NV
 Saint Paul,Minnesota,MN
 St. Louis,Missouri,MO
 Cincinnati,Ohio,OH
 Pittsburgh,Pennsylvania,PA
 Greensboro,North Carolina,NC
 Anchorage,Alaska,AK
 Plano,Texas,TX
 Lincoln,Nebraska,NE
 Orlando,Florida,FL
 Irvine,California,CA
 Newark,New Jersey,NJ
 Durham,North Carolina,NC
 Chula Vista,California,CA
 Toledo,Ohio,OH
 Fort Wayne,Indiana,IN
 St. Petersburg,Florida,FL
 Laredo,Texas,TX
 Jersey City,New Jersey,NJ
 Chandler,Arizona,AZ
 Madison,Wisconsin,WI
 Lubbock,Texas,TX
 Scottsdale,Arizona,AZ
 Reno,Nevada,NV
 Buffalo,New York,NY
 Gilbert,Arizona,AZ
 Glendale,Arizona,AZ
 North Las Vegas,Nevada,NV
 Winston-Salem,North Carolina,NC
 Chesapeake,Virginia,VA
 Norfolk,Virginia,VA
 Fremont,California,CA
 Garland,Texas,TX
 Irving,Texas,TX
 Hialeah,Florida,FL
 Richmond,Virginia,VA
 Boise,Idaho,ID
 Spokane,Washington,WA
 Baton Rouge,Louisiana,LA
@@ -0,0 +1,190 @@
 city,country,country_code
 Toronto,Canada,CA
 Vancouver,Canada,CA
 Montreal,Canada,CA
 Calgary,Canada,CA
 Ottawa,Canada,CA
 London,United Kingdom,GB
 Manchester,United Kingdom,GB
 Birmingham,United Kingdom,GB
 Edinburgh,United Kingdom,GB
 Glasgow,United Kingdom,GB
 Berlin,Germany,DE
 Munich,Germany,DE
 Hamburg,Germany,DE
 Frankfurt,Germany,DE
 Cologne,Germany,DE
 Paris,France,FR
 Lyon,France,FR
 Marseille,France,FR
 Toulouse,France,FR
 Nice,France,FR
 Rome,Italy,IT
 Milan,Italy,IT
 Naples,Italy,IT
 Turin,Italy,IT
 Florence,Italy,IT
 Madrid,Spain,ES
 Barcelona,Spain,ES
 Valencia,Spain,ES
 Seville,Spain,ES
 Bilbao,Spain,ES
 Sydney,Australia,AU
 Melbourne,Australia,AU
 Brisbane,Australia,AU
 Perth,Australia,AU
 Adelaide,Australia,AU
 Tokyo,Japan,JP
 Osaka,Japan,JP
 Kyoto,Japan,JP
 Yokohama,Japan,JP
 Nagoya,Japan,JP
 Beijing,China,CN
 Shanghai,China,CN
 Guangzhou,China,CN
 Shenzhen,China,CN
 Chengdu,China,CN
 Mumbai,India,IN
 Delhi,India,IN
 Bangalore,India,IN
 Hyderabad,India,IN
 Chennai,India,IN
 Sao Paulo,Brazil,BR
 Rio de Janeiro,Brazil,BR
 Brasilia,Brazil,BR
 Salvador,Brazil,BR
 Fortaleza,Brazil,BR
 Mexico City,Mexico,MX
 Guadalajara,Mexico,MX
 Monterrey,Mexico,MX
 Puebla,Mexico,MX
 Tijuana,Mexico,MX
 Moscow,Russia,RU
 Saint Petersburg,Russia,RU
 Novosibirsk,Russia,RU
 Yekaterinburg,Russia,RU
 Kazan,Russia,RU
 Lagos,Nigeria,NG
 Kano,Nigeria,NG
 Ibadan,Nigeria,NG
 Abuja,Nigeria,NG
 Port Harcourt,Nigeria,NG
 Karachi,Pakistan,PK
 Lahore,Pakistan,PK
 Islamabad,Pakistan,PK
 Rawalpindi,Pakistan,PK
 Faisalabad,Pakistan,PK
 Tehran,Iran,IR
 Mashhad,Iran,IR
 Isfahan,Iran,IR
 Karaj,Iran,IR
 Tabriz,Iran,IR
 Pyongyang,North Korea,KP
 Hamhung,North Korea,KP
 Chongjin,North Korea,KP
 Nampo,North Korea,KP
 Wonsan,North Korea,KP
 Damascus,Syria,SY
 Aleppo,Syria,SY
 Homs,Syria,SY
 Latakia,Syria,SY
 Hama,Syria,SY
 Caracas,Venezuela,VE
 Maracaibo,Venezuela,VE
 Valencia,Venezuela,VE
 Barquisimeto,Venezuela,VE
 Maracay,Venezuela,VE
 Havana,Cuba,CU
 Santiago de Cuba,Cuba,CU
 Camaguey,Cuba,CU
 Holguin,Cuba,CU
 Santa Clara,Cuba,CU
 Yangon,Myanmar,MM
 Mandalay,Myanmar,MM
 Naypyidaw,Myanmar,MM
 Mawlamyine,Myanmar,MM
 Bago,Myanmar,MM
 Kabul,Afghanistan,AF
 Kandahar,Afghanistan,AF
 Herat,Afghanistan,AF
 Mazar-i-Sharif,Afghanistan,AF
 Jalalabad,Afghanistan,AF
 Baghdad,Iraq,IQ
 Basra,Iraq,IQ
 Mosul,Iraq,IQ
 Erbil,Iraq,IQ
 Kirkuk,Iraq,IQ
 Tripoli,Libya,LY
 Benghazi,Libya,LY
 Misrata,Libya,LY
 Zawiya,Libya,LY
 Bayda,Libya,LY
 Khartoum,Sudan,SD
 Omdurman,Sudan,SD
 Port Sudan,Sudan,SD
 Kassala,Sudan,SD
 Nyala,Sudan,SD
 Mogadishu,Somalia,SO
 Hargeisa,Somalia,SO
 Bosaso,Somalia,SO
 Kismayo,Somalia,SO
 Merca,Somalia,SO
 Sanaa,Yemen,YE
 Aden,Yemen,YE
 Taiz,Yemen,YE
 Hodeidah,Yemen,YE
 Ibb,Yemen,YE
 Harare,Zimbabwe,ZW
 Bulawayo,Zimbabwe,ZW
 Chitungwiza,Zimbabwe,ZW
 Mutare,Zimbabwe,ZW
 Gweru,Zimbabwe,ZW
 Amsterdam,Netherlands,NL
 Rotterdam,Netherlands,NL
 The Hague,Netherlands,NL
 Utrecht,Netherlands,NL
 Eindhoven,Netherlands,NL
 Stockholm,Sweden,SE
 Gothenburg,Sweden,SE
 Malmo,Sweden,SE
 Uppsala,Sweden,SE
 Vasteras,Sweden,SE
 Oslo,Norway,NO
 Bergen,Norway,NO
 Trondheim,Norway,NO
 Stavanger,Norway,NO
 Drammen,Norway,NO
 Copenhagen,Denmark,DK
 Aarhus,Denmark,DK
 Odense,Denmark,DK
 Aalborg,Denmark,DK
 Esbjerg,Denmark,DK
 Helsinki,Finland,FI
 Espoo,Finland,FI
 Tampere,Finland,FI
 Vantaa,Finland,FI
 Oulu,Finland,FI
 Zurich,Switzerland,CH
 Geneva,Switzerland,CH
 Basel,Switzerland,CH
 Lausanne,Switzerland,CH
 Bern,Switzerland,CH
 Singapore,Singapore,SG
 Hong Kong,Hong Kong,HK
 Kowloon,Hong Kong,HK
 Seoul,South Korea,KR
 Busan,South Korea,KR
 Incheon,South Korea,KR
 Daegu,South Korea,KR
 Daejeon,South Korea,KR
 Taipei,Taiwan,TW
 Kaohsiung,Taiwan,TW
 Taichung,Taiwan,TW
 Tainan,Taiwan,TW
 Hsinchu,Taiwan,TW
 Auckland,New Zealand,NZ
 Wellington,New Zealand,NZ
 Christchurch,New Zealand,NZ
 Hamilton,New Zealand,NZ
 Tauranga,New Zealand,NZ
@@ -0,0 +1,52 @@
 version: '3.8'
 services:
  postgres:
    image: postgres:16-alpine
    container_name: fraud_detection_db
    environment:
      POSTGRES_DB: fraud_detection
      POSTGRES_USER: fraud_analyst
      POSTGRES_PASSWORD: SecurePass123!
      PGDATA: /var/lib/postgresql/data/pgdata
    ports:
      - "5432:5432"
    volumes:
      - postgres_data:/var/lib/postgresql/data
      - ./schema:/docker-entrypoint-initdb.d
      - ./data:/data
    networks:
      - fraud_network
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U fraud_analyst -d fraud_detection"]
      interval: 10s
      timeout: 5s
      retries: 5
    restart: unless-stopped
  db-ui:
    image: ghcr.io/n7olkachev/db-ui:latest
    container_name: fraud_detection_ui
    environment:
      POSTGRES_HOST: postgres
      POSTGRES_USER: fraud_analyst
      POSTGRES_PASSWORD: SecurePass123!
      POSTGRES_DB: fraud_detection
      POSTGRES_PORT: 5432
    ports:
      - "3000:3000"
    networks:
      - fraud_network
    depends_on:
      postgres:
        condition: service_healthy
    restart: unless-stopped
 networks:
  fraud_network:
    driver: bridge
 volumes:
  postgres_data:
    driver: local
@@ -0,0 +1,22 @@
 -- ============================================================================
 -- Database Initialization Script
 -- ============================================================================
 -- This script ensures the database is created and ready
 -- It runs automatically when the PostgreSQL container starts
 -- ============================================================================
 -- Ensure the database exists (this runs in the default postgres database)
 SELECT 'Database initialization starting...' AS status;
 -- Set timezone
 SET timezone = 'UTC';
 -- Show current database
 SELECT current_database() AS current_db, current_user AS current_user, version() AS pg_version;
 -- Enable required extensions
 CREATE EXTENSION IF NOT EXISTS "uuid-ossp";
 CREATE EXTENSION IF NOT EXISTS "pgcrypto";
 SELECT 'Extensions enabled successfully' AS status;
@@ -0,0 +1,313 @@
 # Quick Start Guide
 ## 🚀 Get Up and Running in 5 Minutes
 ### Step 1: Start Docker Containers (1 minute)
 ```bash
 # From the project root directory
 docker-compose up -d
 ```
 **What this does:**
 - Starts PostgreSQL 16 database
 - Starts DB-UI web interface
 - Creates network and volumes
 **Verify it's running:**
 ```bash
 docker-compose ps
 ```
 You should see both `fraud_detection_db` and `fraud_detection_ui` running.
 ---
 ### Step 2: Initialize Database Schema (1 minute)
 ```bash
 # Make script executable (first time only)
 chmod +x scripts/setup-database.sh
 # Run setup
 ./scripts/setup-database.sh
 ```
 **What this does:**
 - Creates all 20+ tables
 - Sets up indexes and constraints
 - Loads reference data (countries, merchant categories, etc.)
 **Expected output:**
 ```
 ✓ PostgreSQL is ready
 ✓ Creating tables, indexes, and constraints
 ✓ Loading reference data
 ✓ Database setup completed successfully!
 ```
 ---
 ### Step 3: Generate Test Data (15-30 minutes)
 ```bash
 # Make script executable (first time only)
 chmod +x data/generate_data.sh
 # Run data generation
 ./data/generate_data.sh
 ```
 **What this does:**
 - Generates 100,000 customers
 - Creates 150,000 accounts
 - Generates 5,000,000 transactions
 - Creates fraud patterns and alerts
 **⏱️ Time estimate:**
 - Fast machine (SSD, 16GB RAM): ~15 minutes
 - Average machine: ~20-25 minutes
 - Slower machine: ~30 minutes
 **You can monitor progress:**
 The script shows progress for each step:
 ```
 [1/9] Loading geographic reference data...
 [2/9] Generating Customers...
 [3/9] Generating Accounts...
 ...
 ```
 ---
 ### Step 4: Access the Database
 #### Option A: DB-UI Web Interface (Recommended for Beginners)
 1. Open your browser to: **http://localhost:3000**
 2. You'll see the database tables in the sidebar
 3. Click any table to browse data
 4. Use the "Custom SQL" tab to run queries
 **Features:**
 - Visual table browser
 - SQL query editor with syntax highlighting
 - Export results to CSV
 - Schema introspection
 #### Option B: Command Line (psql)
 ```bash
 docker exec -it fraud_detection_db psql -U fraud_analyst -d fraud_detection
 ```
 **Quick commands:**
 ```sql
 -- List all tables
 \dt
 -- Describe a table
 \d customers
 -- Run a query
 SELECT COUNT(*) FROM transactions;
 -- Exit
 \q
 ```
 #### Option C: Your Favorite SQL Client
 **Connection Details:**
 ```
 Host:     localhost
 Port:     5432
 Database: fraud_detection
 Username: fraud_analyst
 Password: SecurePass123!
 ```
 **Popular clients:**
 - DBeaver (free, cross-platform)
 - pgAdmin (free, PostgreSQL-specific)
 - DataGrip (paid, JetBrains)
 - TablePlus (paid, macOS/Windows)
 ---
 ## 🎓 Your First Queries
 ### 1. Check Data Counts
 ```sql
 -- How many customers?
 SELECT COUNT(*) FROM customers;
 -- How many transactions?
 SELECT COUNT(*) FROM transactions;
 -- How many fraud alerts?
 SELECT COUNT(*) FROM alerts WHERE status = 'OPEN';
 ```
 ### 2. Find High-Risk Customers
 ```sql
 SELECT 
    customer_id,
    first_name,
    last_name,
    email,
    risk_score
 FROM customers
 WHERE risk_score > 80
 ORDER BY risk_score DESC
 LIMIT 10;
 ```
 ### 3. View Recent Transactions
 ```sql
 SELECT 
    transaction_id,
    account_id,
    amount,
    transaction_date,
    is_flagged
 FROM transactions
 ORDER BY transaction_date DESC
 LIMIT 20;
 ```
 ### 4. Find Flagged Transactions
 ```sql
 SELECT 
    t.transaction_id,
    t.amount,
    t.fraud_score,
    t.flagged_reason,
    c.first_name,
    c.last_name
 FROM transactions t
 JOIN accounts a ON t.account_id = a.account_id
 JOIN customers c ON a.customer_id = c.customer_id
 WHERE t.is_flagged = TRUE
 ORDER BY t.fraud_score DESC
 LIMIT 10;
 ```
 ---
 ## 📚 Next Steps
 ### Start Learning SQL
 1. **Begin with basics:** `exercises/01-basic-queries/README.md`
 2. **Progress through levels:** Work through exercises 01-06
 3. **Practice fraud detection:** `exercises/06-fraud-detection/README.md`
 ### Explore the Data
 ```sql
 -- What countries are represented?
 SELECT country_name, COUNT(*) as customer_count
 FROM customers c
 JOIN countries co ON c.country_id = co.country_id
 GROUP BY country_name
 ORDER BY customer_count DESC;
 -- What are the top merchant categories?
 SELECT mc.category_name, COUNT(*) as transaction_count
 FROM transactions t
 JOIN merchants m ON t.merchant_id = m.merchant_id
 JOIN merchant_categories mc ON m.category_id = mc.category_id
 GROUP BY mc.category_name
 ORDER BY transaction_count DESC;
 -- How many fraud cases by type?
 SELECT ft.fraud_name, COUNT(*) as case_count
 FROM fraud_cases fc
 JOIN fraud_types ft ON fc.fraud_type_id = ft.fraud_type_id
 GROUP BY ft.fraud_name
 ORDER BY case_count DESC;
 ```
 ---
 ## 🔧 Troubleshooting
 ### Database won't start
 ```bash
 # Check logs
 docker-compose logs postgres
 # Restart containers
 docker-compose restart
 ```
 ### Can't connect to database
 ```bash
 # Check if PostgreSQL is ready
 docker exec fraud_detection_db pg_isready -U fraud_analyst
 # Check port is not in use
 netstat -an | grep 5432
 ```
 ### Data generation fails
 ```bash
 # Check disk space
 df -h
 # Check memory
 free -h
 # Try with smaller dataset
 # Edit data/generate_data.sh and reduce:
 NUM_CUSTOMERS=10000
 NUM_TRANSACTIONS=500000
 ```
 ### Reset everything
 ```bash
 # Stop and remove everything
 docker-compose down -v
 # Start fresh
 docker-compose up -d
 ./scripts/setup-database.sh
 ./data/generate_data.sh
 ```
 ---
 ## 💡 Tips
 1. **Use DB-UI for exploration** - Great for browsing and understanding the schema
 2. **Use psql for practice** - Best for learning SQL commands
 3. **Start simple** - Begin with basic SELECT queries before complex joins
 4. **Check the exercises** - They're designed to build your skills progressively
 5. **Experiment** - The database is yours to explore and learn from!
 ---
 ## 🎯 Learning Goals
 After completing this tutorial, you'll be able to:
 - ✅ Write complex SQL queries
 - ✅ Understand database relationships
 - ✅ Detect fraud patterns in data
 - ✅ Use window functions and CTEs
 - ✅ Optimize queries with indexes
 - ✅ Investigate financial crimes
 ---
 **Ready to start? Head to `exercises/01-basic-queries/README.md`!** 🚀
@@ -0,0 +1,206 @@
 # Level 1: Basic SQL Queries
 ## Introduction
 Welcome to the Financial Fraud Detection SQL learning path! In this first level, you'll learn the fundamentals of SQL by querying a realistic fraud detection database.
 ## Learning Objectives
 - Understand SELECT statements
 - Use WHERE clauses for filtering
 - Sort results with ORDER BY
 - Limit result sets
 - Work with basic comparison operators
 ## Exercises
 ### Exercise 1.1: View All Customers
 **Objective:** Retrieve all customer records
 ```sql
 -- Your query here
 SELECT * FROM customers;
 ```
 **Expected Result:** All customer records with all columns
 ---
 ### Exercise 1.2: Find a Specific Customer
 **Objective:** Find customer with customer_id = 1
 ```sql
 -- Your query here
 SELECT * FROM customers WHERE customer_id = 1;
 ```
 ---
 ### Exercise 1.3: High-Risk Customers
 **Objective:** Find all customers with a risk_score greater than 80
 ```sql
 -- Your query here
 ```
 **Hint:** Use the WHERE clause with the > operator
 **Solution:**
 ```sql
 SELECT customer_id, first_name, last_name, email, risk_score
 FROM customers
 WHERE risk_score > 80
 ORDER BY risk_score DESC;
 ```
 ---
 ### Exercise 1.4: Recent Registrations
 **Objective:** Find customers who registered in 2024
 ```sql
 -- Your query here
 ```
 **Hint:** Use WHERE with date comparison
 **Solution:**
 ```sql
 SELECT customer_id, first_name, last_name, email, registration_date
 FROM customers
 WHERE registration_date >= '2024-01-01'
 ORDER BY registration_date DESC;
 ```
 ---
 ### Exercise 1.5: Inactive Accounts
 **Objective:** Find all inactive customer accounts
 ```sql
 -- Your query here
 ```
 **Solution:**
 ```sql
 SELECT customer_id, first_name, last_name, email, is_active
 FROM customers
 WHERE is_active = FALSE;
 ```
 ---
 ### Exercise 1.6: Top 10 Largest Transactions
 **Objective:** Find the 10 largest transactions by amount
 ```sql
 -- Your query here
 ```
 **Hint:** Use ORDER BY with LIMIT
 **Solution:**
 ```sql
 SELECT transaction_id, account_id, amount, transaction_date, description
 FROM transactions
 ORDER BY amount DESC
 LIMIT 10;
 ```
 ---
 ### Exercise 1.7: Flagged Transactions
 **Objective:** Find all transactions that have been flagged for review
 ```sql
 -- Your query here
 ```
 **Solution:**
 ```sql
 SELECT transaction_id, account_id, amount, fraud_score, flagged_reason
 FROM transactions
 WHERE is_flagged = TRUE
 ORDER BY fraud_score DESC;
 ```
 ---
 ### Exercise 1.8: International Transactions
 **Objective:** Find all international transactions over $1,000
 ```sql
 -- Your query here
 ```
 **Solution:**
 ```sql
 SELECT transaction_id, account_id, amount, country_id, city
 FROM transactions
 WHERE is_international = TRUE AND amount > 1000
 ORDER BY amount DESC;
 ```
 ---
 ### Exercise 1.9: Specific Merchant Categories
 **Objective:** Find all merchants in the 'Gambling' or 'Cryptocurrency' categories
 ```sql
 -- Your query here
 ```
 **Hint:** Join merchants with merchant_categories, use IN or OR
 **Solution:**
 ```sql
 SELECT m.merchant_id, m.merchant_name, mc.category_name, m.risk_rating
 FROM merchants m
 JOIN merchant_categories mc ON m.category_id = mc.category_id
 WHERE mc.category_name IN ('Gambling', 'Cryptocurrency')
 ORDER BY m.risk_rating DESC;
 ```
 ---
 ### Exercise 1.10: Critical Alerts
 **Objective:** Find all open alerts with CRITICAL severity
 ```sql
 -- Your query here
 ```
 **Solution:**
 ```sql
 SELECT alert_id, customer_id, alert_type, description, alert_date
 FROM alerts
 WHERE severity = 'CRITICAL' AND status = 'OPEN'
 ORDER BY alert_date DESC;
 ```
 ---
 ## Challenge Exercises
 ### Challenge 1.1: PEP Customers
 Find all Politically Exposed Persons (PEPs) with high risk scores (> 70)
 ### Challenge 1.2: Expired Cards
 Find all cards that have expired (expiry_date < current_date)
 ### Challenge 1.3: Large Cash Advances
 Find all cash advance transactions over $5,000
 ---
 ## Next Steps
 Once you're comfortable with these basic queries, move on to:
 - **Level 2:** JOIN operations
 - **Level 3:** Aggregate functions and GROUP BY
@@ -0,0 +1,360 @@
 # Level 6: Fraud Detection Scenarios
 ## Introduction
 Now you'll apply your SQL skills to real-world fraud detection scenarios. These exercises simulate actual fraud investigation tasks.
 ## Learning Objectives
 - Detect velocity fraud patterns
 - Identify geographic anomalies
 - Find money mule networks
 - Detect account takeover attempts
 - Identify structuring patterns
 ---
 ## Fraud Pattern Detection
 ### Scenario 1: Velocity Fraud Detection
 **Objective:** Find accounts with more than 5 transactions in a 1-hour window
 ```sql
 -- Detect rapid-fire transactions (velocity check)
 WITH transaction_windows AS (
    SELECT 
        t1.account_id,
        t1.transaction_id,
        t1.transaction_date,
        t1.amount,
        COUNT(t2.transaction_id) as transactions_in_hour
    FROM transactions t1
    JOIN transactions t2 ON t1.account_id = t2.account_id
        AND t2.transaction_date BETWEEN t1.transaction_date - INTERVAL '1 hour' 
        AND t1.transaction_date
    GROUP BY t1.account_id, t1.transaction_id, t1.transaction_date, t1.amount
 )
 SELECT 
    account_id,
    transaction_date,
    transactions_in_hour,
    SUM(amount) as total_amount
 FROM transaction_windows
 WHERE transactions_in_hour > 5
 GROUP BY account_id, transaction_date, transactions_in_hour
 ORDER BY transactions_in_hour DESC;
 ```
 ---
 ### Scenario 2: Geographic Impossibility
 **Objective:** Find transactions from the same card in different countries within 2 hours
 ```sql
 -- Detect impossible travel (card used in different countries too quickly)
 SELECT 
    t1.card_id,
    t1.transaction_id as trans1_id,
    t1.transaction_date as trans1_date,
    c1.country_name as country1,
    t1.city as city1,
    t2.transaction_id as trans2_id,
    t2.transaction_date as trans2_date,
    c2.country_name as country2,
    t2.city as city2,
    EXTRACT(EPOCH FROM (t2.transaction_date - t1.transaction_date))/3600 as hours_between
 FROM transactions t1
 JOIN transactions t2 ON t1.card_id = t2.card_id
    AND t2.transaction_date > t1.transaction_date
    AND t2.transaction_date <= t1.transaction_date + INTERVAL '2 hours'
 JOIN countries c1 ON t1.country_id = c1.country_id
 JOIN countries c2 ON t2.country_id = c2.country_id
 WHERE t1.country_id != t2.country_id
 ORDER BY hours_between ASC;
 ```
 ---
 ### Scenario 3: Money Mule Network Detection
 **Objective:** Find clusters of accounts that transfer money in a chain pattern
 ```sql
 -- Detect potential money mule networks (rapid transfer chains)
 WITH transfer_chains AS (
    SELECT 
        tr1.from_account_id as account1,
        tr1.to_account_id as account2,
        tr2.to_account_id as account3,
        tr1.transfer_id as transfer1,
        tr2.transfer_id as transfer2,
        t1.amount as amount1,
        t2.amount as amount2,
        t1.transaction_date as date1,
        t2.transaction_date as date2,
        EXTRACT(EPOCH FROM (t2.transaction_date - t1.transaction_date))/3600 as hours_between
    FROM transfers tr1
    JOIN transfers tr2 ON tr1.to_account_id = tr2.from_account_id
    JOIN transactions t1 ON tr1.transaction_id = t1.transaction_id
    JOIN transactions t2 ON tr2.transaction_id = t2.transaction_id
    WHERE t2.transaction_date BETWEEN t1.transaction_date AND t1.transaction_date + INTERVAL '24 hours'
 )
 SELECT 
    account1,
    account2,
    account3,
    amount1,
    amount2,
    hours_between,
    CASE 
        WHEN ABS(amount1 - amount2) / amount1 < 0.1 THEN 'SUSPICIOUS - Similar amounts'
        ELSE 'Review'
    END as risk_flag
 FROM transfer_chains
 WHERE hours_between < 24
 ORDER BY hours_between ASC;
 ```
 ---
 ### Scenario 4: Account Takeover Detection
 **Objective:** Find accounts with sudden changes in transaction patterns
 ```sql
 -- Detect account takeover by analyzing behavior changes
 WITH customer_baseline AS (
    SELECT 
        a.customer_id,
        a.account_id,
        AVG(t.amount) as avg_transaction,
        STDDEV(t.amount) as stddev_transaction,
        COUNT(*) as transaction_count
    FROM accounts a
    JOIN transactions t ON a.account_id = t.account_id
    WHERE t.transaction_date < CURRENT_DATE - INTERVAL '30 days'
    GROUP BY a.customer_id, a.account_id
 ),
 recent_transactions AS (
    SELECT 
        a.customer_id,
        a.account_id,
        t.transaction_id,
        t.amount,
        t.transaction_date,
        t.country_id,
        t.device_id
    FROM accounts a
    JOIN transactions t ON a.account_id = t.account_id
    WHERE t.transaction_date >= CURRENT_DATE - INTERVAL '7 days'
 )
 SELECT 
    rt.customer_id,
    rt.account_id,
    rt.transaction_id,
    rt.amount,
    cb.avg_transaction,
    (rt.amount - cb.avg_transaction) / NULLIF(cb.stddev_transaction, 0) as z_score,
    CASE 
        WHEN ABS((rt.amount - cb.avg_transaction) / NULLIF(cb.stddev_transaction, 0)) > 3 
        THEN 'HIGH RISK - Amount anomaly'
        WHEN ABS((rt.amount - cb.avg_transaction) / NULLIF(cb.stddev_transaction, 0)) > 2 
        THEN 'MEDIUM RISK'
        ELSE 'Normal'
    END as risk_level
 FROM recent_transactions rt
 JOIN customer_baseline cb ON rt.account_id = cb.account_id
 WHERE cb.transaction_count > 10
 ORDER BY ABS((rt.amount - cb.avg_transaction) / NULLIF(cb.stddev_transaction, 0)) DESC;
 ```
 ---
 ### Scenario 5: Structuring Detection (Smurfing)
 **Objective:** Find patterns of transactions just under $10,000 (reporting threshold)
 ```sql
 -- Detect structuring - multiple transactions just under reporting threshold
 WITH daily_transactions AS (
    SELECT 
        account_id,
        DATE(transaction_date) as transaction_day,
        COUNT(*) as num_transactions,
        SUM(amount) as total_amount,
        AVG(amount) as avg_amount,
        MAX(amount) as max_amount
    FROM transactions
    WHERE amount BETWEEN 9000 AND 9999
        AND transaction_date >= CURRENT_DATE - INTERVAL '30 days'
    GROUP BY account_id, DATE(transaction_date)
 )
 SELECT 
    dt.account_id,
    c.first_name,
    c.last_name,
    dt.transaction_day,
    dt.num_transactions,
    dt.total_amount,
    dt.avg_amount,
    CASE 
        WHEN dt.num_transactions >= 3 AND dt.total_amount > 25000 
        THEN 'CRITICAL - Likely structuring'
        WHEN dt.num_transactions >= 2 AND dt.total_amount > 18000 
        THEN 'HIGH - Possible structuring'
        ELSE 'Review'
    END as risk_assessment
 FROM daily_transactions dt
 JOIN accounts a ON dt.account_id = a.account_id
 JOIN customers c ON a.customer_id = c.customer_id
 WHERE dt.num_transactions >= 2
 ORDER BY dt.total_amount DESC, dt.num_transactions DESC;
 ```
 ---
 ### Scenario 6: High-Risk Merchant Analysis
 **Objective:** Find customers with unusual activity at high-risk merchants
 ```sql
 -- Analyze transactions at high-risk merchants
 SELECT 
    c.customer_id,
    c.first_name,
    c.last_name,
    c.risk_score as customer_risk,
    mc.category_name,
    m.merchant_name,
    m.risk_rating as merchant_risk,
    COUNT(t.transaction_id) as transaction_count,
    SUM(t.amount) as total_spent,
    AVG(t.amount) as avg_transaction,
    MAX(t.amount) as max_transaction
 FROM customers c
 JOIN accounts a ON c.customer_id = a.customer_id
 JOIN transactions t ON a.account_id = t.account_id
 JOIN merchants m ON t.merchant_id = m.merchant_id
 JOIN merchant_categories mc ON m.category_id = mc.category_id
 WHERE m.risk_rating IN ('HIGH', 'CRITICAL')
    AND t.transaction_date >= CURRENT_DATE - INTERVAL '90 days'
 GROUP BY c.customer_id, c.first_name, c.last_name, c.risk_score, 
         mc.category_name, m.merchant_name, m.risk_rating
 HAVING COUNT(t.transaction_id) > 5 OR SUM(t.amount) > 10000
 ORDER BY total_spent DESC;
 ```
 ---
 ### Scenario 7: Card Testing Detection
 **Objective:** Find cards with multiple small failed transactions (testing stolen cards)
 ```sql
 -- Detect card testing patterns
 SELECT 
    t.card_id,
    c.card_last_four,
    COUNT(*) as failed_attempts,
    COUNT(DISTINCT t.merchant_id) as different_merchants,
    MIN(t.amount) as min_amount,
    MAX(t.amount) as max_amount,
    MIN(t.transaction_date) as first_attempt,
    MAX(t.transaction_date) as last_attempt,
    EXTRACT(EPOCH FROM (MAX(t.transaction_date) - MIN(t.transaction_date)))/60 as minutes_span
 FROM transactions t
 JOIN cards c ON t.card_id = c.card_id
 WHERE t.status = 'FAILED'
    AND t.amount < 10
    AND t.transaction_date >= CURRENT_DATE - INTERVAL '24 hours'
 GROUP BY t.card_id, c.card_last_four
 HAVING COUNT(*) >= 3
 ORDER BY failed_attempts DESC, minutes_span ASC;
 ```
 ---
 ### Scenario 8: Dormant Account Reactivation
 **Objective:** Find dormant accounts that suddenly become active (potential takeover)
 ```sql
 -- Detect dormant account reactivation
 WITH account_activity AS (
    SELECT 
        account_id,
        MIN(transaction_date) as first_transaction,
        MAX(transaction_date) as last_transaction,
        COUNT(*) as total_transactions
    FROM transactions
    GROUP BY account_id
 ),
 dormant_accounts AS (
    SELECT 
        account_id,
        last_transaction,
        total_transactions
    FROM account_activity
    WHERE last_transaction < CURRENT_DATE - INTERVAL '180 days'
 ),
 recent_activity AS (
    SELECT 
        t.account_id,
        COUNT(*) as recent_transactions,
        SUM(t.amount) as recent_amount,
        MIN(t.transaction_date) as reactivation_date
    FROM transactions t
    WHERE t.transaction_date >= CURRENT_DATE - INTERVAL '7 days'
    GROUP BY t.account_id
 )
 SELECT 
    da.account_id,
    c.first_name,
    c.last_name,
    c.email,
    da.last_transaction as last_active,
    EXTRACT(DAY FROM (CURRENT_DATE - da.last_transaction)) as days_dormant,
    ra.reactivation_date,
    ra.recent_transactions,
    ra.recent_amount,
    'CRITICAL - Dormant account reactivated' as alert_type
 FROM dormant_accounts da
 JOIN recent_activity ra ON da.account_id = ra.account_id
 JOIN accounts a ON da.account_id = a.account_id
 JOIN customers c ON a.customer_id = c.customer_id
 ORDER BY days_dormant DESC;
 ```
 ---
 ## Investigation Exercises
 ### Exercise 6.1: Full Customer Investigation
 Create a comprehensive report for a suspicious customer including:
 - All accounts
 - All transactions
 - All alerts
 - All fraud cases
 - Related customers (via relationships)
 ### Exercise 6.2: Fraud Case Summary
 Generate a summary report of all open fraud cases with:
 - Case details
 - Associated transactions
 - Total amount at risk
 - Investigation status
 ### Exercise 6.3: Daily Fraud Dashboard
 Create a daily dashboard showing:
 - New alerts by severity
 - High-risk transactions
 - Geographic anomalies
 - Velocity violations
 ---
 ## Next Steps
 Congratulations! You've completed the fraud detection scenarios. You now have the skills to:
 - Detect complex fraud patterns
 - Investigate suspicious activity
 - Generate fraud reports
 - Analyze customer behavior
 Continue practicing with real data and explore advanced topics like:
 - Machine learning integration
 - Real-time fraud scoring
 - Network analysis
 - Predictive modeling
@@ -0,0 +1,477 @@
 -- ============================================================================
 -- Financial Fraud Detection Database Schema
 -- Purpose: Educational SQL learning with realistic fraud investigation scenarios
 -- ============================================================================
 -- This script is IDEMPOTENT - it will drop and recreate all objects
 -- ============================================================================
 -- Enable required extensions
 CREATE EXTENSION IF NOT EXISTS "uuid-ossp";
 CREATE EXTENSION IF NOT EXISTS "pgcrypto";
 -- ============================================================================
 -- DROP ALL EXISTING OBJECTS (in reverse dependency order)
 -- ============================================================================
 -- Drop triggers first
 DROP TRIGGER IF EXISTS trg_update_balance ON transactions;
 DROP TRIGGER IF EXISTS trg_check_suspicious ON transactions;
 -- Drop functions
 DROP FUNCTION IF EXISTS update_account_balance() CASCADE;
 DROP FUNCTION IF EXISTS check_suspicious_transaction() CASCADE;
 DROP FUNCTION IF EXISTS generate_fraud_score(DECIMAL, BOOLEAN, DECIMAL, INT, BOOLEAN) CASCADE;
 -- Drop tables in reverse dependency order
 DROP TABLE IF EXISTS audit_log CASCADE;
 DROP TABLE IF EXISTS suspicious_activity_reports CASCADE;
 DROP TABLE IF EXISTS case_alerts CASCADE;
 DROP TABLE IF EXISTS case_transactions CASCADE;
 DROP TABLE IF EXISTS fraud_cases CASCADE;
 DROP TABLE IF EXISTS alerts CASCADE;
 DROP TABLE IF EXISTS transfers CASCADE;
 DROP TABLE IF EXISTS beneficiaries CASCADE;
 DROP TABLE IF EXISTS transactions CASCADE;
 DROP TABLE IF EXISTS login_sessions CASCADE;
 DROP TABLE IF EXISTS devices CASCADE;
 DROP TABLE IF EXISTS cards CASCADE;
 DROP TABLE IF EXISTS accounts CASCADE;
 DROP TABLE IF EXISTS customer_relationships CASCADE;
 DROP TABLE IF EXISTS customers CASCADE;
 DROP TABLE IF EXISTS merchants CASCADE;
 DROP TABLE IF EXISTS merchant_categories CASCADE;
 DROP TABLE IF EXISTS fraud_types CASCADE;
 DROP TABLE IF EXISTS transaction_types CASCADE;
 DROP TABLE IF EXISTS countries CASCADE;
 -- ============================================================================
 -- REFERENCE/LOOKUP TABLES
 -- ============================================================================
 -- Countries reference table
 CREATE TABLE countries (
    country_id SERIAL PRIMARY KEY,
    country_code CHAR(2) NOT NULL UNIQUE,
    country_name VARCHAR(100) NOT NULL,
    region VARCHAR(50) NOT NULL,
    risk_level VARCHAR(20) DEFAULT 'LOW' CHECK (risk_level IN ('LOW', 'MEDIUM', 'HIGH', 'CRITICAL')),
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Merchant categories
 CREATE TABLE merchant_categories (
    category_id SERIAL PRIMARY KEY,
    category_code VARCHAR(10) NOT NULL UNIQUE,
    category_name VARCHAR(100) NOT NULL,
    description TEXT,
    risk_weight DECIMAL(3,2) DEFAULT 1.00,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Transaction types
 CREATE TABLE transaction_types (
    type_id SERIAL PRIMARY KEY,
    type_code VARCHAR(20) NOT NULL UNIQUE,
    type_name VARCHAR(100) NOT NULL,
    description TEXT,
    requires_merchant BOOLEAN DEFAULT FALSE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Fraud types
 CREATE TABLE fraud_types (
    fraud_type_id SERIAL PRIMARY KEY,
    fraud_code VARCHAR(20) NOT NULL UNIQUE,
    fraud_name VARCHAR(100) NOT NULL,
    description TEXT,
    severity VARCHAR(20) DEFAULT 'MEDIUM' CHECK (severity IN ('LOW', 'MEDIUM', 'HIGH', 'CRITICAL')),
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- ============================================================================
 -- CUSTOMER DOMAIN
 -- ============================================================================
 -- Customers table
 CREATE TABLE customers (
    customer_id BIGSERIAL PRIMARY KEY,
    first_name VARCHAR(100) NOT NULL,
    last_name VARCHAR(100) NOT NULL,
    email VARCHAR(255) NOT NULL UNIQUE,
    phone VARCHAR(20),
    date_of_birth DATE NOT NULL,
    ssn_hash VARCHAR(64) NOT NULL UNIQUE, -- Hashed SSN for privacy
    address_line1 VARCHAR(255),
    address_line2 VARCHAR(255),
    city VARCHAR(100),
    state VARCHAR(50),
    postal_code VARCHAR(20),
    country_id INT NOT NULL REFERENCES countries(country_id),
    registration_date TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
    last_login TIMESTAMP,
    kyc_status VARCHAR(20) DEFAULT 'PENDING' CHECK (kyc_status IN ('PENDING', 'VERIFIED', 'REJECTED', 'EXPIRED')),
    kyc_verified_date TIMESTAMP,
    risk_score DECIMAL(5,2) DEFAULT 50.00 CHECK (risk_score BETWEEN 0 AND 100),
    is_pep BOOLEAN DEFAULT FALSE, -- Politically Exposed Person
    is_active BOOLEAN DEFAULT TRUE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Customer relationships (for detecting collusion networks)
 CREATE TABLE customer_relationships (
    relationship_id BIGSERIAL PRIMARY KEY,
    customer_id_1 BIGINT NOT NULL REFERENCES customers(customer_id),
    customer_id_2 BIGINT NOT NULL REFERENCES customers(customer_id),
    relationship_type VARCHAR(50) NOT NULL CHECK (relationship_type IN ('FAMILY', 'BUSINESS', 'SHARED_ADDRESS', 'SHARED_DEVICE', 'SHARED_IP', 'SUSPECTED_MULE')),
    confidence_score DECIMAL(5,2) DEFAULT 50.00 CHECK (confidence_score BETWEEN 0 AND 100),
    detected_date TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    notes TEXT,
    CONSTRAINT different_customers CHECK (customer_id_1 != customer_id_2),
    CONSTRAINT unique_relationship UNIQUE (customer_id_1, customer_id_2, relationship_type)
 );
 -- ============================================================================
 -- ACCOUNT DOMAIN
 -- ============================================================================
 -- Accounts table
 CREATE TABLE accounts (
    account_id BIGSERIAL PRIMARY KEY,
    customer_id BIGINT NOT NULL REFERENCES customers(customer_id),
    account_number VARCHAR(20) NOT NULL UNIQUE,
    account_type VARCHAR(20) NOT NULL CHECK (account_type IN ('CHECKING', 'SAVINGS', 'CREDIT', 'INVESTMENT', 'LOAN')),
    currency CHAR(3) DEFAULT 'USD',
    opening_date DATE NOT NULL DEFAULT CURRENT_DATE,
    closing_date DATE,
    status VARCHAR(20) DEFAULT 'ACTIVE' CHECK (status IN ('ACTIVE', 'SUSPENDED', 'CLOSED', 'FROZEN')),
    current_balance DECIMAL(15,2) DEFAULT 0.00,
    available_balance DECIMAL(15,2) DEFAULT 0.00,
    credit_limit DECIMAL(15,2),
    overdraft_limit DECIMAL(15,2) DEFAULT 0.00,
    interest_rate DECIMAL(5,4),
    monthly_fee DECIMAL(8,2) DEFAULT 0.00,
    is_primary BOOLEAN DEFAULT FALSE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Cards table
 CREATE TABLE cards (
    card_id BIGSERIAL PRIMARY KEY,
    account_id BIGINT NOT NULL REFERENCES accounts(account_id),
    card_number_hash VARCHAR(64) NOT NULL UNIQUE, -- Hashed card number
    card_last_four CHAR(4) NOT NULL,
    card_type VARCHAR(20) NOT NULL CHECK (card_type IN ('DEBIT', 'CREDIT', 'PREPAID', 'VIRTUAL')),
    card_network VARCHAR(20) NOT NULL CHECK (card_network IN ('VISA', 'MASTERCARD', 'AMEX', 'DISCOVER')),
    issue_date DATE NOT NULL DEFAULT CURRENT_DATE,
    expiry_date DATE NOT NULL,
    cvv_hash VARCHAR(64) NOT NULL,
    status VARCHAR(20) DEFAULT 'ACTIVE' CHECK (status IN ('ACTIVE', 'BLOCKED', 'EXPIRED', 'LOST', 'STOLEN')),
    daily_limit DECIMAL(10,2) DEFAULT 5000.00,
    monthly_limit DECIMAL(12,2) DEFAULT 50000.00,
    is_contactless BOOLEAN DEFAULT TRUE,
    is_international BOOLEAN DEFAULT FALSE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- ============================================================================
 -- MERCHANT DOMAIN
 -- ============================================================================
 -- Merchants table
 CREATE TABLE merchants (
    merchant_id BIGSERIAL PRIMARY KEY,
    merchant_name VARCHAR(255) NOT NULL,
    merchant_code VARCHAR(50) UNIQUE,
    category_id INT NOT NULL REFERENCES merchant_categories(category_id),
    country_id INT NOT NULL REFERENCES countries(country_id),
    city VARCHAR(100),
    website VARCHAR(255),
    registration_date DATE NOT NULL DEFAULT CURRENT_DATE,
    status VARCHAR(20) DEFAULT 'ACTIVE' CHECK (status IN ('ACTIVE', 'SUSPENDED', 'BLACKLISTED', 'CLOSED')),
    risk_rating VARCHAR(20) DEFAULT 'LOW' CHECK (risk_rating IN ('LOW', 'MEDIUM', 'HIGH', 'CRITICAL')),
    total_transactions BIGINT DEFAULT 0,
    total_volume DECIMAL(18,2) DEFAULT 0.00,
    fraud_incidents INT DEFAULT 0,
    is_verified BOOLEAN DEFAULT FALSE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- ============================================================================
 -- DEVICE & SESSION DOMAIN
 -- ============================================================================
 -- Devices table (for tracking login devices)
 CREATE TABLE devices (
    device_id BIGSERIAL PRIMARY KEY,
    device_fingerprint VARCHAR(64) NOT NULL UNIQUE,
    device_type VARCHAR(20) CHECK (device_type IN ('MOBILE', 'TABLET', 'DESKTOP', 'OTHER')),
    os_name VARCHAR(50),
    os_version VARCHAR(50),
    browser_name VARCHAR(50),
    browser_version VARCHAR(50),
    first_seen TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    last_seen TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    is_trusted BOOLEAN DEFAULT FALSE,
    is_blacklisted BOOLEAN DEFAULT FALSE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Login sessions
 CREATE TABLE login_sessions (
    session_id BIGSERIAL PRIMARY KEY,
    customer_id BIGINT NOT NULL REFERENCES customers(customer_id),
    device_id BIGINT NOT NULL REFERENCES devices(device_id),
    ip_address INET NOT NULL,
    country_id INT REFERENCES countries(country_id),
    city VARCHAR(100),
    latitude DECIMAL(10,8),
    longitude DECIMAL(11,8),
    login_timestamp TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    logout_timestamp TIMESTAMP,
    session_duration_seconds INT,
    is_successful BOOLEAN DEFAULT TRUE,
    failure_reason VARCHAR(255),
    risk_score DECIMAL(5,2) DEFAULT 0.00,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- ============================================================================
 -- TRANSACTION DOMAIN
 -- ============================================================================
 -- Transactions table (main transaction log)
 CREATE TABLE transactions (
    transaction_id BIGSERIAL PRIMARY KEY,
    account_id BIGINT NOT NULL REFERENCES accounts(account_id),
    type_id INT NOT NULL REFERENCES transaction_types(type_id),
    transaction_date TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
    amount DECIMAL(15,2) NOT NULL CHECK (amount > 0),
    currency CHAR(3) DEFAULT 'USD',
    merchant_id BIGINT REFERENCES merchants(merchant_id),
    card_id BIGINT REFERENCES cards(card_id),
    device_id BIGINT REFERENCES devices(device_id),
    ip_address INET,
    country_id INT REFERENCES countries(country_id),
    city VARCHAR(100),
    latitude DECIMAL(10,8),
    longitude DECIMAL(11,8),
    description TEXT,
    reference_number VARCHAR(50) UNIQUE,
    status VARCHAR(20) DEFAULT 'COMPLETED' CHECK (status IN ('PENDING', 'COMPLETED', 'FAILED', 'REVERSED', 'FLAGGED', 'BLOCKED')),
    is_online BOOLEAN DEFAULT TRUE,
    is_international BOOLEAN DEFAULT FALSE,
    is_card_present BOOLEAN DEFAULT FALSE,
    fraud_score DECIMAL(5,2) DEFAULT 0.00 CHECK (fraud_score BETWEEN 0 AND 100),
    is_flagged BOOLEAN DEFAULT FALSE,
    flagged_reason TEXT,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Beneficiaries (for transfers)
 CREATE TABLE beneficiaries (
    beneficiary_id BIGSERIAL PRIMARY KEY,
    customer_id BIGINT NOT NULL REFERENCES customers(customer_id),
    beneficiary_name VARCHAR(255) NOT NULL,
    account_number VARCHAR(50) NOT NULL,
    bank_name VARCHAR(255),
    bank_code VARCHAR(20),
    country_id INT NOT NULL REFERENCES countries(country_id),
    relationship VARCHAR(50),
    is_verified BOOLEAN DEFAULT FALSE,
    added_date TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    last_used TIMESTAMP,
    total_transfers INT DEFAULT 0,
    total_amount DECIMAL(18,2) DEFAULT 0.00,
    is_active BOOLEAN DEFAULT TRUE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Transfer transactions
 CREATE TABLE transfers (
    transfer_id BIGSERIAL PRIMARY KEY,
    transaction_id BIGINT NOT NULL REFERENCES transactions(transaction_id),
    from_account_id BIGINT NOT NULL REFERENCES accounts(account_id),
    to_account_id BIGINT REFERENCES accounts(account_id), -- NULL for external transfers
    beneficiary_id BIGINT REFERENCES beneficiaries(beneficiary_id),
    transfer_type VARCHAR(20) NOT NULL CHECK (transfer_type IN ('INTERNAL', 'DOMESTIC', 'INTERNATIONAL', 'WIRE')),
    purpose VARCHAR(255),
    is_recurring BOOLEAN DEFAULT FALSE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- ============================================================================
 -- FRAUD DETECTION & ALERTS DOMAIN
 -- ============================================================================
 -- Alerts table (system-generated suspicious activity alerts)
 CREATE TABLE alerts (
    alert_id BIGSERIAL PRIMARY KEY,
    transaction_id BIGINT REFERENCES transactions(transaction_id),
    customer_id BIGINT REFERENCES customers(customer_id),
    account_id BIGINT REFERENCES accounts(account_id),
    alert_type VARCHAR(50) NOT NULL CHECK (alert_type IN (
        'VELOCITY_CHECK', 'AMOUNT_ANOMALY', 'GEOGRAPHIC_ANOMALY',
        'MERCHANT_RISK', 'DEVICE_CHANGE', 'UNUSUAL_TIME',
        'MULTIPLE_CARDS', 'ACCOUNT_TAKEOVER', 'MONEY_MULE', 'STRUCTURING'
    )),
    severity VARCHAR(20) DEFAULT 'MEDIUM' CHECK (severity IN ('LOW', 'MEDIUM', 'HIGH', 'CRITICAL')),
    alert_date TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    description TEXT NOT NULL,
    risk_score DECIMAL(5,2) DEFAULT 50.00 CHECK (risk_score BETWEEN 0 AND 100),
    status VARCHAR(20) DEFAULT 'OPEN' CHECK (status IN ('OPEN', 'INVESTIGATING', 'CLOSED', 'FALSE_POSITIVE', 'CONFIRMED_FRAUD')),
    assigned_to VARCHAR(100),
    reviewed_date TIMESTAMP,
    resolution_notes TEXT,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Fraud cases (confirmed fraud incidents)
 CREATE TABLE fraud_cases (
    case_id BIGSERIAL PRIMARY KEY,
    case_number VARCHAR(50) NOT NULL UNIQUE,
    customer_id BIGINT REFERENCES customers(customer_id),
    account_id BIGINT REFERENCES accounts(account_id),
    fraud_type_id INT NOT NULL REFERENCES fraud_types(fraud_type_id),
    detection_date TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
    detection_method VARCHAR(50) CHECK (detection_method IN ('AUTOMATED', 'CUSTOMER_REPORT', 'MANUAL_REVIEW', 'THIRD_PARTY')),
    amount_lost DECIMAL(15,2) DEFAULT 0.00,
    amount_recovered DECIMAL(15,2) DEFAULT 0.00,
    status VARCHAR(20) DEFAULT 'OPEN' CHECK (status IN ('OPEN', 'INVESTIGATING', 'RESOLVED', 'CLOSED', 'LEGAL_ACTION')),
    priority VARCHAR(20) DEFAULT 'MEDIUM' CHECK (priority IN ('LOW', 'MEDIUM', 'HIGH', 'CRITICAL')),
    assigned_investigator VARCHAR(100),
    investigation_notes TEXT,
    resolution_date TIMESTAMP,
    resolution_summary TEXT,
    law_enforcement_notified BOOLEAN DEFAULT FALSE,
    customer_notified BOOLEAN DEFAULT FALSE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- Case transactions (linking transactions to fraud cases)
 CREATE TABLE case_transactions (
    case_transaction_id BIGSERIAL PRIMARY KEY,
    case_id BIGINT NOT NULL REFERENCES fraud_cases(case_id),
    transaction_id BIGINT NOT NULL REFERENCES transactions(transaction_id),
    is_fraudulent BOOLEAN DEFAULT TRUE,
    notes TEXT,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    CONSTRAINT unique_case_transaction UNIQUE (case_id, transaction_id)
 );
 -- Case alerts (linking alerts to fraud cases)
 CREATE TABLE case_alerts (
    case_alert_id BIGSERIAL PRIMARY KEY,
    case_id BIGINT NOT NULL REFERENCES fraud_cases(case_id),
    alert_id BIGINT NOT NULL REFERENCES alerts(alert_id),
    relevance_score DECIMAL(5,2) DEFAULT 50.00,
    notes TEXT,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    CONSTRAINT unique_case_alert UNIQUE (case_id, alert_id)
 );
 -- ============================================================================
 -- AUDIT & COMPLIANCE DOMAIN
 -- ============================================================================
 -- Audit log (comprehensive audit trail)
 CREATE TABLE audit_log (
    audit_id BIGSERIAL PRIMARY KEY,
    table_name VARCHAR(100) NOT NULL,
    record_id BIGINT NOT NULL,
    action VARCHAR(20) NOT NULL CHECK (action IN ('INSERT', 'UPDATE', 'DELETE', 'SELECT')),
    old_values JSONB,
    new_values JSONB,
    changed_by VARCHAR(100),
    changed_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    ip_address INET,
    user_agent TEXT
 );
 -- Suspicious Activity Reports (SAR)
 CREATE TABLE suspicious_activity_reports (
    sar_id BIGSERIAL PRIMARY KEY,
    sar_number VARCHAR(50) NOT NULL UNIQUE,
    case_id BIGINT REFERENCES fraud_cases(case_id),
    customer_id BIGINT NOT NULL REFERENCES customers(customer_id),
    filing_date DATE NOT NULL DEFAULT CURRENT_DATE,
    activity_date_from DATE NOT NULL,
    activity_date_to DATE NOT NULL,
    total_amount DECIMAL(18,2) NOT NULL,
    activity_description TEXT NOT NULL,
    filed_by VARCHAR(100) NOT NULL,
    status VARCHAR(20) DEFAULT 'DRAFT' CHECK (status IN ('DRAFT', 'SUBMITTED', 'ACKNOWLEDGED', 'CLOSED')),
    submission_date DATE,
    acknowledgment_date DATE,
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
 );
 -- ============================================================================
 -- INDEXES FOR PERFORMANCE
 -- ============================================================================
 -- Customer indexes
 CREATE INDEX idx_customers_email ON customers(email);
 CREATE INDEX idx_customers_country ON customers(country_id);
 CREATE INDEX idx_customers_risk_score ON customers(risk_score DESC);
 CREATE INDEX idx_customers_registration_date ON customers(registration_date);
 -- Account indexes
 CREATE INDEX idx_accounts_customer ON accounts(customer_id);
 CREATE INDEX idx_accounts_status ON accounts(status);
 CREATE INDEX idx_accounts_type ON accounts(account_type);
 -- Transaction indexes (critical for performance)
 CREATE INDEX idx_transactions_account ON transactions(account_id);
 CREATE INDEX idx_transactions_date ON transactions(transaction_date DESC);
 CREATE INDEX idx_transactions_merchant ON transactions(merchant_id);
 CREATE INDEX idx_transactions_status ON transactions(status);
 CREATE INDEX idx_transactions_flagged ON transactions(is_flagged) WHERE is_flagged = TRUE;
 CREATE INDEX idx_transactions_fraud_score ON transactions(fraud_score DESC);
 CREATE INDEX idx_transactions_amount ON transactions(amount);
 CREATE INDEX idx_transactions_country ON transactions(country_id);
 -- Card indexes
 CREATE INDEX idx_cards_account ON cards(account_id);
 CREATE INDEX idx_cards_status ON cards(status);
 -- Alert indexes
 CREATE INDEX idx_alerts_customer ON alerts(customer_id);
 CREATE INDEX idx_alerts_transaction ON alerts(transaction_id);
 CREATE INDEX idx_alerts_status ON alerts(status);
 CREATE INDEX idx_alerts_date ON alerts(alert_date DESC);
 CREATE INDEX idx_alerts_severity ON alerts(severity);
 -- Fraud case indexes
 CREATE INDEX idx_fraud_cases_customer ON fraud_cases(customer_id);
 CREATE INDEX idx_fraud_cases_status ON fraud_cases(status);
 CREATE INDEX idx_fraud_cases_detection_date ON fraud_cases(detection_date DESC);
 -- Login session indexes
 CREATE INDEX idx_login_sessions_customer ON login_sessions(customer_id);
 CREATE INDEX idx_login_sessions_timestamp ON login_sessions(login_timestamp DESC);
 CREATE INDEX idx_login_sessions_ip ON login_sessions(ip_address);
 -- Merchant indexes
 CREATE INDEX idx_merchants_category ON merchants(category_id);
 CREATE INDEX idx_merchants_country ON merchants(country_id);
 CREATE INDEX idx_merchants_risk_rating ON merchants(risk_rating);
 -- Comments for documentation
 COMMENT ON TABLE customers IS 'Customer master data with KYC and risk information';
 COMMENT ON TABLE accounts IS 'Customer accounts including checking, savings, credit, etc.';
 COMMENT ON TABLE transactions IS 'Main transaction log with fraud scoring';
 COMMENT ON TABLE alerts IS 'System-generated fraud alerts requiring review';
 COMMENT ON TABLE fraud_cases IS 'Confirmed fraud cases under investigation';
 COMMENT ON TABLE merchants IS 'Merchant directory with risk ratings';
 COMMENT ON TABLE cards IS 'Payment cards linked to accounts';
 COMMENT ON TABLE devices IS 'Device fingerprints for fraud detection';
 COMMENT ON TABLE login_sessions IS 'Login history for account takeover detection';
@@ -0,0 +1,256 @@
 -- ============================================================================
 -- Seed Data for Reference Tables
 -- ============================================================================
 -- This script is IDEMPOTENT - it will delete and recreate all reference data
 -- ============================================================================
 -- Clear existing reference data (in reverse dependency order)
 TRUNCATE TABLE suspicious_activity_reports CASCADE;
 TRUNCATE TABLE case_alerts CASCADE;
 TRUNCATE TABLE case_transactions CASCADE;
 TRUNCATE TABLE fraud_cases CASCADE;
 TRUNCATE TABLE alerts CASCADE;
 TRUNCATE TABLE transfers CASCADE;
 TRUNCATE TABLE beneficiaries CASCADE;
 TRUNCATE TABLE transactions CASCADE;
 TRUNCATE TABLE login_sessions CASCADE;
 TRUNCATE TABLE devices CASCADE;
 TRUNCATE TABLE cards CASCADE;
 TRUNCATE TABLE accounts CASCADE;
 TRUNCATE TABLE customer_relationships CASCADE;
 TRUNCATE TABLE customers CASCADE;
 TRUNCATE TABLE merchants CASCADE;
 TRUNCATE TABLE merchant_categories RESTART IDENTITY CASCADE;
 TRUNCATE TABLE fraud_types RESTART IDENTITY CASCADE;
 TRUNCATE TABLE transaction_types RESTART IDENTITY CASCADE;
 TRUNCATE TABLE countries RESTART IDENTITY CASCADE;
 -- Insert Countries
 INSERT INTO countries (country_code, country_name, region, risk_level) VALUES
 ('US', 'United States', 'North America', 'LOW'),
 ('CA', 'Canada', 'North America', 'LOW'),
 ('GB', 'United Kingdom', 'Europe', 'LOW'),
 ('DE', 'Germany', 'Europe', 'LOW'),
 ('FR', 'France', 'Europe', 'LOW'),
 ('IT', 'Italy', 'Europe', 'MEDIUM'),
 ('ES', 'Spain', 'Europe', 'MEDIUM'),
 ('AU', 'Australia', 'Oceania', 'LOW'),
 ('JP', 'Japan', 'Asia', 'LOW'),
 ('CN', 'China', 'Asia', 'MEDIUM'),
 ('IN', 'India', 'Asia', 'MEDIUM'),
 ('BR', 'Brazil', 'South America', 'MEDIUM'),
 ('MX', 'Mexico', 'North America', 'MEDIUM'),
 ('RU', 'Russia', 'Europe', 'HIGH'),
 ('NG', 'Nigeria', 'Africa', 'HIGH'),
 ('PK', 'Pakistan', 'Asia', 'HIGH'),
 ('IR', 'Iran', 'Middle East', 'CRITICAL'),
 ('KP', 'North Korea', 'Asia', 'CRITICAL'),
 ('SY', 'Syria', 'Middle East', 'CRITICAL'),
 ('VE', 'Venezuela', 'South America', 'HIGH'),
 ('CU', 'Cuba', 'Caribbean', 'HIGH'),
 ('MM', 'Myanmar', 'Asia', 'HIGH'),
 ('AF', 'Afghanistan', 'Asia', 'CRITICAL'),
 ('IQ', 'Iraq', 'Middle East', 'HIGH'),
 ('LY', 'Libya', 'Africa', 'HIGH'),
 ('SD', 'Sudan', 'Africa', 'HIGH'),
 ('SO', 'Somalia', 'Africa', 'CRITICAL'),
 ('YE', 'Yemen', 'Middle East', 'CRITICAL'),
 ('ZW', 'Zimbabwe', 'Africa', 'HIGH'),
 ('NL', 'Netherlands', 'Europe', 'LOW'),
 ('SE', 'Sweden', 'Europe', 'LOW'),
 ('NO', 'Norway', 'Europe', 'LOW'),
 ('DK', 'Denmark', 'Europe', 'LOW'),
 ('FI', 'Finland', 'Europe', 'LOW'),
 ('CH', 'Switzerland', 'Europe', 'LOW'),
 ('SG', 'Singapore', 'Asia', 'LOW'),
 ('HK', 'Hong Kong', 'Asia', 'MEDIUM'),
 ('KR', 'South Korea', 'Asia', 'LOW'),
 ('TW', 'Taiwan', 'Asia', 'LOW'),
 ('NZ', 'New Zealand', 'Oceania', 'LOW');
 -- Insert Merchant Categories (based on MCC codes)
 INSERT INTO merchant_categories (category_code, category_name, description, risk_weight) VALUES
 ('5411', 'Grocery Stores', 'Supermarkets and grocery stores', 0.50),
 ('5812', 'Restaurants', 'Eating places and restaurants', 0.60),
 ('5541', 'Gas Stations', 'Service stations and fuel', 0.55),
 ('5311', 'Department Stores', 'General merchandise stores', 0.70),
 ('5912', 'Pharmacies', 'Drug stores and pharmacies', 0.50),
 ('5999', 'Miscellaneous Retail', 'Specialty retail stores', 0.80),
 ('5732', 'Electronics', 'Electronics and computer stores', 1.20),
 ('5651', 'Clothing', 'Family clothing stores', 0.75),
 ('5814', 'Fast Food', 'Quick service restaurants', 0.60),
 ('5942', 'Books', 'Book stores', 0.65),
 ('5945', 'Hobby Shops', 'Hobby, toy, and game shops', 0.70),
 ('5971', 'Art Dealers', 'Art dealers and galleries', 1.50),
 ('5993', 'Cigar Stores', 'Cigar stores and stands', 1.10),
 ('5995', 'Pet Shops', 'Pet shops and supplies', 0.70),
 ('7011', 'Hotels', 'Lodging and hotels', 0.90),
 ('7512', 'Car Rental', 'Automobile rental agencies', 1.00),
 ('7523', 'Parking', 'Parking lots and garages', 0.60),
 ('7832', 'Movie Theaters', 'Motion picture theaters', 0.65),
 ('7922', 'Theatrical Producers', 'Theatrical producers and ticket agencies', 0.80),
 ('7991', 'Tourist Attractions', 'Tourist attractions and exhibits', 0.75),
 ('7995', 'Gambling', 'Betting and casino gambling', 2.50),
 ('5816', 'Digital Goods', 'Digital goods and games', 1.80),
 ('5967', 'Direct Marketing', 'Direct marketing and inbound telemarketing', 1.90),
 ('5966', 'Direct Marketing', 'Outbound telemarketing merchants', 2.00),
 ('6051', 'Cryptocurrency', 'Cryptocurrency and digital currency', 3.00),
 ('6211', 'Securities', 'Securities brokers and dealers', 1.50),
 ('6300', 'Insurance', 'Insurance sales and underwriting', 1.20),
 ('6513', 'Real Estate', 'Real estate agents and managers', 1.30),
 ('7273', 'Dating Services', 'Dating and escort services', 2.20),
 ('7297', 'Massage Parlors', 'Massage parlors', 2.50),
 ('7995', 'Online Gambling', 'Online gambling and betting', 3.50),
 ('5094', 'Precious Metals', 'Precious stones and metals', 2.80),
 ('5933', 'Pawn Shops', 'Pawn shops', 2.60),
 ('5960', 'Mail Order', 'Direct marketing and mail order', 1.70),
 ('4829', 'Wire Transfer', 'Money transfer services', 2.40);
 -- Insert Transaction Types
 INSERT INTO transaction_types (type_code, type_name, description, requires_merchant) VALUES
 ('PURCHASE', 'Purchase', 'Card purchase at merchant', TRUE),
 ('ATM_WITHDRAWAL', 'ATM Withdrawal', 'Cash withdrawal from ATM', FALSE),
 ('DEPOSIT', 'Deposit', 'Cash or check deposit', FALSE),
 ('TRANSFER_OUT', 'Transfer Out', 'Outgoing transfer', FALSE),
 ('TRANSFER_IN', 'Transfer In', 'Incoming transfer', FALSE),
 ('PAYMENT', 'Bill Payment', 'Bill payment transaction', TRUE),
 ('REFUND', 'Refund', 'Merchant refund', TRUE),
 ('FEE', 'Fee', 'Bank fee or charge', FALSE),
 ('INTEREST', 'Interest', 'Interest credit', FALSE),
 ('WIRE_OUT', 'Wire Transfer Out', 'Outgoing wire transfer', FALSE),
 ('WIRE_IN', 'Wire Transfer In', 'Incoming wire transfer', FALSE),
 ('CHECK', 'Check Payment', 'Check payment', FALSE),
 ('DIRECT_DEBIT', 'Direct Debit', 'Automated direct debit', TRUE),
 ('CASH_ADVANCE', 'Cash Advance', 'Credit card cash advance', FALSE),
 ('BALANCE_TRANSFER', 'Balance Transfer', 'Credit card balance transfer', FALSE);
 -- Insert Fraud Types
 INSERT INTO fraud_types (fraud_code, fraud_name, description, severity) VALUES
 ('CARD_NOT_PRESENT', 'Card Not Present Fraud', 'Fraudulent online or phone transactions', 'HIGH'),
 ('CARD_STOLEN', 'Stolen Card', 'Transactions using stolen physical card', 'HIGH'),
 ('ACCOUNT_TAKEOVER', 'Account Takeover', 'Unauthorized access to customer account', 'CRITICAL'),
 ('IDENTITY_THEFT', 'Identity Theft', 'Fraudulent account opened with stolen identity', 'CRITICAL'),
 ('FRIENDLY_FRAUD', 'Friendly Fraud', 'Customer disputes legitimate transaction', 'MEDIUM'),
 ('MONEY_MULE', 'Money Mule', 'Account used to launder money', 'CRITICAL'),
 ('SYNTHETIC_IDENTITY', 'Synthetic Identity', 'Fake identity using real and fake information', 'CRITICAL'),
 ('BUST_OUT', 'Bust Out Fraud', 'Building credit then maxing out and disappearing', 'HIGH'),
 ('REFUND_FRAUD', 'Refund Fraud', 'Fraudulent refund requests', 'MEDIUM'),
 ('CHARGEBACK_FRAUD', 'Chargeback Fraud', 'Abusing chargeback process', 'MEDIUM'),
 ('ATM_SKIMMING', 'ATM Skimming', 'Card data stolen via ATM skimmer', 'HIGH'),
 ('PHISHING', 'Phishing', 'Credentials stolen via phishing attack', 'HIGH'),
 ('SIM_SWAP', 'SIM Swap', 'Phone number hijacked for 2FA bypass', 'CRITICAL'),
 ('CHECK_FRAUD', 'Check Fraud', 'Fraudulent or altered checks', 'MEDIUM'),
 ('WIRE_FRAUD', 'Wire Fraud', 'Fraudulent wire transfer', 'CRITICAL'),
 ('STRUCTURING', 'Structuring', 'Breaking up transactions to avoid reporting', 'HIGH'),
 ('SMURFING', 'Smurfing', 'Using multiple people to structure transactions', 'HIGH'),
 ('TRADE_BASED', 'Trade-Based Money Laundering', 'Using trade to launder money', 'CRITICAL'),
 ('SHELL_COMPANY', 'Shell Company', 'Using fake companies for fraud', 'CRITICAL'),
 ('INVOICE_FRAUD', 'Invoice Fraud', 'Fraudulent invoicing schemes', 'HIGH');
 -- Create a function to generate realistic transaction patterns
 CREATE OR REPLACE FUNCTION generate_fraud_score(
    p_amount DECIMAL,
    p_is_international BOOLEAN,
    p_merchant_risk DECIMAL,
    p_time_of_day INT,
    p_is_online BOOLEAN
 ) RETURNS DECIMAL AS $$
 DECLARE
    v_score DECIMAL := 0;
 BEGIN
    -- Amount-based scoring
    IF p_amount > 5000 THEN v_score := v_score + 20; END IF;
    IF p_amount > 10000 THEN v_score := v_score + 30; END IF;
    -- International transactions
    IF p_is_international THEN v_score := v_score + 15; END IF;
    -- Merchant risk
    v_score := v_score + (p_merchant_risk * 10);
    -- Time of day (late night transactions)
    IF p_time_of_day >= 23 OR p_time_of_day <= 4 THEN v_score := v_score + 10; END IF;
    -- Online transactions
    IF p_is_online THEN v_score := v_score + 5; END IF;
    -- Cap at 100
    IF v_score > 100 THEN v_score := 100; END IF;
    RETURN v_score;
 END;
 $$ LANGUAGE plpgsql;
 -- Create a function to update account balances
 CREATE OR REPLACE FUNCTION update_account_balance()
 RETURNS TRIGGER AS $$
 BEGIN
    IF NEW.status = 'COMPLETED' THEN
        IF TG_TABLE_NAME = 'transactions' THEN
            -- Update based on transaction type
            UPDATE accounts 
            SET current_balance = current_balance + 
                CASE 
                    WHEN NEW.type_id IN (SELECT type_id FROM transaction_types WHERE type_code IN ('DEPOSIT', 'TRANSFER_IN', 'WIRE_IN', 'REFUND', 'INTEREST')) 
                    THEN NEW.amount
                    ELSE -NEW.amount
                END,
                updated_at = CURRENT_TIMESTAMP
            WHERE account_id = NEW.account_id;
        END IF;
    END IF;
    RETURN NEW;
 END;
 $$ LANGUAGE plpgsql;
 -- Create trigger for balance updates (commented out for bulk loading)
 -- CREATE TRIGGER trg_update_balance
 -- AFTER INSERT ON transactions
 -- FOR EACH ROW
 -- EXECUTE FUNCTION update_account_balance();
 -- Create a function to auto-generate alerts for suspicious transactions
 CREATE OR REPLACE FUNCTION check_suspicious_transaction()
 RETURNS TRIGGER AS $$
 DECLARE
    v_alert_type VARCHAR(50);
    v_description TEXT;
    v_severity VARCHAR(20);
 BEGIN
    -- High amount transactions
    IF NEW.amount > 10000 THEN
        v_alert_type := 'AMOUNT_ANOMALY';
        v_description := 'Large transaction amount: $' || NEW.amount;
        v_severity := 'HIGH';
        INSERT INTO alerts (transaction_id, customer_id, account_id, alert_type, severity, description, risk_score)
        SELECT NEW.transaction_id, a.customer_id, NEW.account_id, v_alert_type, v_severity, v_description, NEW.fraud_score
        FROM accounts a WHERE a.account_id = NEW.account_id;
    END IF;
    -- International transactions
    IF NEW.is_international AND NEW.amount > 1000 THEN
        v_alert_type := 'GEOGRAPHIC_ANOMALY';
        v_description := 'International transaction: $' || NEW.amount;
        v_severity := 'MEDIUM';
        INSERT INTO alerts (transaction_id, customer_id, account_id, alert_type, severity, description, risk_score)
        SELECT NEW.transaction_id, a.customer_id, NEW.account_id, v_alert_type, v_severity, v_description, NEW.fraud_score
        FROM accounts a WHERE a.account_id = NEW.account_id;
    END IF;
    RETURN NEW;
 END;
 $$ LANGUAGE plpgsql;
 -- Create trigger for alert generation (commented out for bulk loading)
 -- CREATE TRIGGER trg_check_suspicious
 -- AFTER INSERT ON transactions
 -- FOR EACH ROW
 -- WHEN (NEW.fraud_score > 50)
 -- EXECUTE FUNCTION check_suspicious_transaction();
 COMMENT ON FUNCTION generate_fraud_score IS 'Calculates fraud risk score based on transaction attributes';
 COMMENT ON FUNCTION update_account_balance IS 'Automatically updates account balance after transaction';
 COMMENT ON FUNCTION check_suspicious_transaction IS 'Generates alerts for suspicious transactions';
@@ -0,0 +1,114 @@
 #!/bin/bash
 # ============================================================================
 # Database Setup Script - IDEMPOTENT
 # ============================================================================
 # This script sets up the complete fraud detection database
 # It can be run multiple times safely - it will recreate everything
 # ============================================================================
 set -e  # Exit on error
 # Colors for output
 RED='\033[0;31m'
 GREEN='\033[0;32m'
 YELLOW='\033[1;33m'
 BLUE='\033[0;34m'
 NC='\033[0m' # No Color
 # Configuration from environment or defaults
 DB_HOST="${POSTGRES_HOST:-localhost}"
 DB_PORT="${POSTGRES_PORT:-5432}"
 DB_NAME="${POSTGRES_DB:-fraud_detection}"
 DB_USER="${POSTGRES_USER:-fraud_analyst}"
 DB_PASSWORD="${POSTGRES_PASSWORD:-SecurePass123!}"
 # Script directory
 SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
 PROJECT_ROOT="$(dirname "$SCRIPT_DIR")"
 echo -e "${BLUE}============================================================================${NC}"
 echo -e "${BLUE}Financial Fraud Detection Database - Setup Script${NC}"
 echo -e "${BLUE}============================================================================${NC}"
 echo -e "Database: ${GREEN}$DB_NAME${NC}"
 echo -e "Host: ${GREEN}$DB_HOST:$DB_PORT${NC}"
 echo -e "User: ${GREEN}$DB_USER${NC}"
 echo -e "${BLUE}============================================================================${NC}"
 echo ""
 # Function to execute SQL command
 execute_sql() {
    PGPASSWORD=$DB_PASSWORD psql -h $DB_HOST -p $DB_PORT -U $DB_USER -d $DB_NAME -c "$1" 2>&1
 }
 # Function to execute SQL file
 execute_sql_file() {
    local file=$1
    local description=$2
    echo -e "${YELLOW}Executing: $description${NC}"
    PGPASSWORD=$DB_PASSWORD psql -h $DB_HOST -p $DB_PORT -U $DB_USER -d $DB_NAME -f "$file" 2>&1
    if [ $? -eq 0 ]; then
        echo -e "${GREEN}✓ Success: $description${NC}"
    else
        echo -e "${RED}✗ Failed: $description${NC}"
        exit 1
    fi
    echo ""
 }
 # Wait for PostgreSQL to be ready
 echo -e "${YELLOW}Waiting for PostgreSQL to be ready...${NC}"
 max_attempts=30
 attempt=0
 until PGPASSWORD=$DB_PASSWORD psql -h $DB_HOST -p $DB_PORT -U $DB_USER -d $DB_NAME -c '\q' 2>/dev/null; do
    attempt=$((attempt + 1))
    if [ $attempt -ge $max_attempts ]; then
        echo -e "${RED}✗ PostgreSQL is not available after $max_attempts attempts${NC}"
        exit 1
    fi
    echo -e "${YELLOW}Waiting for PostgreSQL... (attempt $attempt/$max_attempts)${NC}"
    sleep 2
 done
 echo -e "${GREEN}✓ PostgreSQL is ready${NC}"
 echo ""
 # Step 1: Create schema (drops and recreates all tables)
 echo -e "${BLUE}[Step 1/3] Creating database schema...${NC}"
 execute_sql_file "$PROJECT_ROOT/schema/01-create-tables.sql" "Creating tables, indexes, and constraints"
 # Step 2: Load seed data (reference tables)
 echo -e "${BLUE}[Step 2/3] Loading reference data...${NC}"
 execute_sql_file "$PROJECT_ROOT/schema/02-seed-data.sql" "Loading countries, merchant categories, transaction types, and fraud types"
 # Step 3: Verify setup
 echo -e "${BLUE}[Step 3/3] Verifying database setup...${NC}"
 echo -e "${YELLOW}Checking table counts...${NC}"
 # Get table counts
 table_count=$(execute_sql "SELECT COUNT(*) FROM information_schema.tables WHERE table_schema = 'public' AND table_type = 'BASE TABLE';" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Tables created: ${GREEN}$table_count${NC}"
 # Get reference data counts
 country_count=$(execute_sql "SELECT COUNT(*) FROM countries;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Countries: ${GREEN}$country_count${NC}"
 category_count=$(execute_sql "SELECT COUNT(*) FROM merchant_categories;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Merchant categories: ${GREEN}$category_count${NC}"
 transaction_type_count=$(execute_sql "SELECT COUNT(*) FROM transaction_types;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Transaction types: ${GREEN}$transaction_type_count${NC}"
 fraud_type_count=$(execute_sql "SELECT COUNT(*) FROM fraud_types;" | grep -E '^\s*[0-9]+' | tr -d ' ')
 echo -e "Fraud types: ${GREEN}$fraud_type_count${NC}"
 echo ""
 echo -e "${GREEN}============================================================================${NC}"
 echo -e "${GREEN}✓ Database setup completed successfully!${NC}"
 echo -e "${GREEN}============================================================================${NC}"
 echo ""
 echo -e "${YELLOW}Next steps:${NC}"
 echo -e "1. Generate test data: ${BLUE}./scripts/generate-data.sh${NC}"
 echo -e "2. Access DB-UI at: ${BLUE}http://localhost:3000${NC}"
 echo -e "3. Start learning SQL with exercises in: ${BLUE}./exercises/${NC}"
 echo ""
@@ -0,0 +1,210 @@
 #!/bin/bash
 # ============================================================================
 # Setup Verification Script
 # ============================================================================
 # Verifies that the fraud detection database is properly set up
 # ============================================================================
 set -e
 # Colors for output
 RED='\033[0;31m'
 GREEN='\033[0;32m'
 YELLOW='\033[1;33m'
 BLUE='\033[0;34m'
 NC='\033[0m' # No Color
 # Configuration
 DB_HOST="${POSTGRES_HOST:-localhost}"
 DB_PORT="${POSTGRES_PORT:-5432}"
 DB_NAME="${POSTGRES_DB:-fraud_detection}"
 DB_USER="${POSTGRES_USER:-fraud_analyst}"
 DB_PASSWORD="${POSTGRES_PASSWORD:-SecurePass123!}"
 echo -e "${BLUE}============================================================================${NC}"
 echo -e "${BLUE}Financial Fraud Detection Database - Verification${NC}"
 echo -e "${BLUE}============================================================================${NC}"
 echo ""
 # Function to execute SQL and get result
 execute_sql() {
    PGPASSWORD=$DB_PASSWORD psql -h $DB_HOST -p $DB_PORT -U $DB_USER -d $DB_NAME -t -c "$1" 2>/dev/null | xargs
 }
 # Check 1: Docker containers
 echo -e "${YELLOW}[1/10] Checking Docker containers...${NC}"
 if docker ps | grep -q "fraud_detection_db"; then
    echo -e "${GREEN}✓ PostgreSQL container is running${NC}"
 else
    echo -e "${RED}✗ PostgreSQL container is not running${NC}"
    echo -e "${YELLOW}Run: docker-compose up -d${NC}"
    exit 1
 fi
 if docker ps | grep -q "fraud_detection_ui"; then
    echo -e "${GREEN}✓ DB-UI container is running${NC}"
 else
    echo -e "${RED}✗ DB-UI container is not running${NC}"
    echo -e "${YELLOW}Run: docker-compose up -d${NC}"
    exit 1
 fi
 echo ""
 # Check 2: Database connectivity
 echo -e "${YELLOW}[2/10] Checking database connectivity...${NC}"
 if PGPASSWORD=$DB_PASSWORD psql -h $DB_HOST -p $DB_PORT -U $DB_USER -d $DB_NAME -c '\q' 2>/dev/null; then
    echo -e "${GREEN}✓ Can connect to database${NC}"
 else
    echo -e "${RED}✗ Cannot connect to database${NC}"
    exit 1
 fi
 echo ""
 # Check 3: Tables exist
 echo -e "${YELLOW}[3/10] Checking database schema...${NC}"
 table_count=$(execute_sql "SELECT COUNT(*) FROM information_schema.tables WHERE table_schema = 'public' AND table_type = 'BASE TABLE';")
 if [ "$table_count" -ge 20 ]; then
    echo -e "${GREEN}✓ Schema created ($table_count tables)${NC}"
 else
    echo -e "${RED}✗ Schema incomplete (only $table_count tables)${NC}"
    echo -e "${YELLOW}Run: ./scripts/setup-database.sh${NC}"
    exit 1
 fi
 echo ""
 # Check 4: Reference data
 echo -e "${YELLOW}[4/10] Checking reference data...${NC}"
 country_count=$(execute_sql "SELECT COUNT(*) FROM countries;")
 category_count=$(execute_sql "SELECT COUNT(*) FROM merchant_categories;")
 type_count=$(execute_sql "SELECT COUNT(*) FROM transaction_types;")
 fraud_type_count=$(execute_sql "SELECT COUNT(*) FROM fraud_types;")
 if [ "$country_count" -ge 40 ]; then
    echo -e "${GREEN}✓ Countries loaded ($country_count)${NC}"
 else
    echo -e "${RED}✗ Countries not loaded${NC}"
 fi
 if [ "$category_count" -ge 30 ]; then
    echo -e "${GREEN}✓ Merchant categories loaded ($category_count)${NC}"
 else
    echo -e "${RED}✗ Merchant categories not loaded${NC}"
 fi
 if [ "$type_count" -ge 10 ]; then
    echo -e "${GREEN}✓ Transaction types loaded ($type_count)${NC}"
 else
    echo -e "${RED}✗ Transaction types not loaded${NC}"
 fi
 if [ "$fraud_type_count" -ge 15 ]; then
    echo -e "${GREEN}✓ Fraud types loaded ($fraud_type_count)${NC}"
 else
    echo -e "${RED}✗ Fraud types not loaded${NC}"
 fi
 echo ""
 # Check 5: Customer data
 echo -e "${YELLOW}[5/10] Checking customer data...${NC}"
 customer_count=$(execute_sql "SELECT COUNT(*) FROM customers;")
 if [ "$customer_count" -ge 10000 ]; then
    echo -e "${GREEN}✓ Customers generated ($customer_count)${NC}"
 else
    echo -e "${YELLOW}⚠ Limited customer data ($customer_count)${NC}"
    echo -e "${YELLOW}Run: ./data/generate_data.sh${NC}"
 fi
 echo ""
 # Check 6: Account data
 echo -e "${YELLOW}[6/10] Checking account data...${NC}"
 account_count=$(execute_sql "SELECT COUNT(*) FROM accounts;")
 if [ "$account_count" -ge 10000 ]; then
    echo -e "${GREEN}✓ Accounts generated ($account_count)${NC}"
 else
    echo -e "${YELLOW}⚠ Limited account data ($account_count)${NC}"
 fi
 echo ""
 # Check 7: Transaction data
 echo -e "${YELLOW}[7/10] Checking transaction data...${NC}"
 transaction_count=$(execute_sql "SELECT COUNT(*) FROM transactions;")
 if [ "$transaction_count" -ge 100000 ]; then
    echo -e "${GREEN}✓ Transactions generated ($transaction_count)${NC}"
 else
    echo -e "${YELLOW}⚠ Limited transaction data ($transaction_count)${NC}"
 fi
 echo ""
 # Check 8: Fraud data
 echo -e "${YELLOW}[8/10] Checking fraud detection data...${NC}"
 alert_count=$(execute_sql "SELECT COUNT(*) FROM alerts;")
 case_count=$(execute_sql "SELECT COUNT(*) FROM fraud_cases;")
 if [ "$alert_count" -ge 100 ]; then
    echo -e "${GREEN}✓ Alerts generated ($alert_count)${NC}"
 else
    echo -e "${YELLOW}⚠ Limited alert data ($alert_count)${NC}"
 fi
 if [ "$case_count" -ge 10 ]; then
    echo -e "${GREEN}✓ Fraud cases generated ($case_count)${NC}"
 else
    echo -e "${YELLOW}⚠ Limited fraud case data ($case_count)${NC}"
 fi
 echo ""
 # Check 9: Indexes
 echo -e "${YELLOW}[9/10] Checking database indexes...${NC}"
 index_count=$(execute_sql "SELECT COUNT(*) FROM pg_indexes WHERE schemaname = 'public';")
 if [ "$index_count" -ge 20 ]; then
    echo -e "${GREEN}✓ Indexes created ($index_count)${NC}"
 else
    echo -e "${YELLOW}⚠ Limited indexes ($index_count)${NC}"
 fi
 echo ""
 # Check 10: DB-UI accessibility
 echo -e "${YELLOW}[10/10] Checking DB-UI web interface...${NC}"
 if curl -s -o /dev/null -w "%{http_code}" http://localhost:3000 | grep -q "200\|302"; then
    echo -e "${GREEN}✓ DB-UI accessible at http://localhost:3000${NC}"
 else
    echo -e "${YELLOW}⚠ DB-UI may not be ready yet (still starting up)${NC}"
 fi
 echo ""
 # Summary
 echo -e "${BLUE}============================================================================${NC}"
 echo -e "${GREEN}Verification Summary${NC}"
 echo -e "${BLUE}============================================================================${NC}"
 echo ""
 echo -e "${YELLOW}Database Statistics:${NC}"
 echo -e "  Customers:     ${GREEN}$customer_count${NC}"
 echo -e "  Accounts:      ${GREEN}$account_count${NC}"
 echo -e "  Transactions:  ${GREEN}$transaction_count${NC}"
 echo -e "  Alerts:        ${GREEN}$alert_count${NC}"
 echo -e "  Fraud Cases:   ${GREEN}$case_count${NC}"
 echo ""
 if [ "$transaction_count" -ge 100000 ]; then
    echo -e "${GREEN}✓ Database is ready for SQL learning!${NC}"
    echo ""
    echo -e "${YELLOW}Next steps:${NC}"
    echo -e "1. Access DB-UI: ${BLUE}http://localhost:3000${NC}"
    echo -e "2. Start learning: ${BLUE}exercises/01-basic-queries/README.md${NC}"
    echo -e "3. Quick start guide: ${BLUE}docs/QUICKSTART.md${NC}"
 else
    echo -e "${YELLOW}⚠ Database has minimal data${NC}"
    echo ""
    echo -e "${YELLOW}To generate full dataset:${NC}"
    echo -e "  ${BLUE}./data/generate_data.sh${NC}"
    echo ""
    echo -e "${YELLOW}This will generate:${NC}"
    echo -e "  - 100,000 customers"
    echo -e "  - 150,000 accounts"
    echo -e "  - 5,000,000 transactions"
    echo -e "  - 50,000+ alerts"
    echo -e "  - 5,000+ fraud cases"
 fi
 echo ""