Financial Crime Monitoring System - Bank of Singapore Prototype

A comprehensive AI-powered financial crime detection and monitoring system built with Google ADK agents, featuring synthetic transaction data generation and real-time risk assessment for Singapore banking operations.

🏦 Overview

This system is specifically designed as a prototype for Bank of Singapore, providing intelligent financial crime monitoring with:

• AI-Powered Analysis: Multi-agent system using Google ADK for sophisticated transaction analysis
• Synthetic Data Generation: Realistic SGD transaction data using Python Faker
• Real-time Monitoring: Live transaction risk assessment and flagging
• Singapore Context: All amounts in SGD, calibrated for Singapore banking market
• Modern Web Interface: React + TypeScript frontend with shadcn/ui components

🏗️ System Architecture

┌─────────────────┐    HTTP/REST API    ┌──────────────────┐
│  React Frontend │ ◄─────────────────► │  FastAPI Backend │
│  (TypeScript)   │                     │   (Python)       │
│                 │                     │                  │
│ • Dashboard     │                     │ • ADK Agents     │
│ • Search/Filter │                     │ • Data Generator │
│ • Transaction   │                     │ • Risk Analysis  │
│   Details       │                     │ • API Endpoints  │
│ • AI Analysis   │                     │                  │
└─────────────────┘                     └──────────────────┘

🚀 Quick Start

Prerequisites

• Node.js (v18+) with pnpm
• Python (v3.11+)
• uv (Python package manager)

1. Clone the Repository

git clone <repository-url>
cd adk-fcc

2. Backend Setup

# Navigate to backend directory
cd backend

# Install Python dependencies
uv pip install -r requirements.txt

# Start the FastAPI server
uv run uvicorn main:app --host 0.0.0.0 --port 8000 --reload

The backend will be available at: http://localhost:8000

3. Frontend Setup

# Navigate back to project root
cd ..

# Install frontend dependencies
pnpm install

# Start the development server
pnpm dev

The frontend will be available at: http://localhost:5173 (or http://localhost:5174 if 5173 is in use)

4. Access the Application

Open your browser and navigate to the frontend URL. The system will automatically connect to the backend and start displaying synthetic transaction data.

📁 Project Structure

adk-fcc/
├── backend/                    # Python FastAPI backend
│   ├── agents/                # Google ADK agents
│   │   ├── __init__.py
│   │   ├── financial_crime_coordinator.py
│   │   ├── prompts.py
│   │   └── sub_agents/        # Specialized sub-agents
│   │       ├── compliance_checker.py
│   │       ├── pattern_detector.py
│   │       ├── risk_assessor.py
│   │       └── transaction_analyzer.py
│   ├── data_generator.py      # Synthetic data generation
│   ├── main.py               # FastAPI application
│   ├── requirements.txt      # Python dependencies
│   └── start.sh             # Startup script
├── src/                      # React frontend
│   ├── components/          # React components
│   │   ├── ui/             # shadcn/ui components
│   │   ├── Dashboard.tsx   # Main dashboard
│   │   ├── TransactionTable.tsx
│   │   ├── TransactionDetails.tsx
│   │   ├── AIAnalysis.tsx
│   │   └── SearchBar.tsx
│   ├── services/           
│   │   └── api.ts          # API service layer
│   ├── types/
│   │   └── transaction.ts  # TypeScript interfaces
│   └── hooks/              # Custom React hooks
├── package.json            # Frontend dependencies
└── README.md              # This file

🎯 Key Features

Backend Features

• 🤖 Multi-Agent AI System: Coordinated agents for comprehensive analysis

  • Transaction Analyzer
  • Risk Assessor
  • Compliance Checker
  • Pattern Detector

• 🎲 Synthetic Data Generation: Realistic SGD transactions with:

  • Risk-based distribution (30% high-risk, 70% normal)
  • Geographic diversity (32+ countries)
  • Realistic amount ranges for Singapore market
  • Customer profiles and merchant information

• 🔍 Risk Analysis: AI-generated analysis including:

  • Multi-factor risk scoring
  • Detailed risk factors and reasoning
  • Compliance screening results
  • Pattern detection indicators

Frontend Features

• 📊 Interactive Dashboard: Real-time transaction monitoring
• 🔍 Advanced Search & Filtering: Filter by risk level, status, transaction ID
• 📋 Transaction Details: Comprehensive transaction information
• 🧠 AI Analysis Viewer: Detailed risk assessment results
• ⚡ Real-time Updates: Live data refresh and status updates

🏃‍♂️ Usage

API Endpoints

Endpoint	Method	Description
/health	GET	Health check
/api/transactions/flagged	GET	Get synthetic transactions
/api/transactions/{id}	GET	Get specific transaction
/api/analysis/{id}	GET	Get AI analysis for transaction
/api/transactions/{id}/status	PATCH	Update transaction status
/api/agents/status	GET	Get agent status

Example API Calls

# Get 10 synthetic transactions
curl "http://localhost:8000/api/transactions/flagged?count=10"

# Get high-risk transactions only
curl "http://localhost:8000/api/transactions/flagged?risk_indicator=High"

# Get AI analysis for a transaction
curl "http://localhost:8000/api/analysis/TXN-12345678"

# Update transaction status
curl -X PATCH "http://localhost:8000/api/transactions/TXN-12345678/status" \
  -H "Content-Type: application/json" \
  -d '{"status": "reviewed"}'

SGD Amount Ranges

The system generates realistic SGD amounts appropriate for Singapore banking:

Transaction Type	Normal Range	High-Risk Range
Transfers	SGD 500 - 50,000	SGD 50,000 - 2,000,000
Deposits	SGD 200 - 25,000	SGD 25,000 - 800,000
Withdrawals	SGD 100 - 15,000	SGD 15,000 - 200,000
Payments	SGD 50 - 5,000	SGD 5,000 - 150,000

🛠️ Development

Backend Development

cd backend

# Run data generator test
uv run python data_generator.py

# Start with auto-reload for development
uv run uvicorn main:app --reload --host 0.0.0.0 --port 8000

Frontend Development

# Start with hot reload
pnpm dev

# Type checking
pnpm type-check

# Linting
pnpm lint

# Build for production
pnpm build

Environment Configuration

Create a .env file in the backend/ directory based on .env.example:

cp backend/.env.example backend/.env

🧪 Testing

Test Backend

cd backend

# Test data generation
uv run python data_generator.py

# Test API endpoints
curl http://localhost:8000/health
curl "http://localhost:8000/api/transactions/flagged?count=5"

Test Frontend

# Install dependencies and start
pnpm install
pnpm dev

# Open browser to http://localhost:5173

🚢 Deployment

Backend Deployment

cd backend

# Install production dependencies
uv pip install -r requirements.txt

# Start production server
uv run uvicorn main:app --host 0.0.0.0 --port 8000

Frontend Deployment

# Build for production
pnpm build

# Preview production build
pnpm preview

# Deploy dist/ folder to your hosting service

🔧 Configuration

Backend Configuration

• Port: Default 8000 (configurable in main.py)
• CORS: Configured for localhost:5173, localhost:3000
• Data Generation: Configurable transaction counts and risk percentages

Frontend Configuration

• API Base URL: http://localhost:8000/api (configurable in src/services/api.ts)
• Timeout: 15 seconds for API calls
• Fallback Data: Available when backend is unreachable

📚 Additional Documentation

• Synthetic Data Implementation: Detailed guide on data generation
• Backend API Documentation: Available at http://localhost:8000/docs when running
• Component Documentation: See individual component files in src/components/

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/new-feature)
3. Commit your changes (git commit -am 'Add new feature')
4. Push to the branch (git push origin feature/new-feature)
5. Create a Pull Request

📄 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

🏦 Bank of Singapore Context

This prototype is specifically designed for Bank of Singapore with:

• SGD Currency: All transactions in Singapore Dollar
• Local Market Calibration: Amount ranges appropriate for Singapore banking
• Regulatory Alignment: Risk factors aligned with MAS (Monetary Authority of Singapore) guidelines
• Regional Risk Assessment: Geographic risk categories relevant to Singapore's position

---

Built with ❤️ for Bank of Singapore

For support or questions, please contact the development team.