Julius Baer: Agentic AI AML Platform

A comprehensive AI-powered platform for real-time Anti-Money Laundering (AML) monitoring and document corroboration, built for Julius Baer.

🎯 Project Overview

This platform integrates two critical AML capabilities:

1. Real-Time AML Monitoring - Monitor transactions in real-time, generate risk-based alerts, and provide actionable insights
2. Document & Image Corroboration - AI-powered document forensics to detect tampering, inconsistencies, and suspicious patterns

✨ Features

Dashboard

• Real-time KPI metrics (Active Alerts, Critical Cases, Resolution Times)
• Risk level distribution visualization
• Transaction volume trends
• Comprehensive alert triage queue

Investigation Cockpit

• Detailed transaction analysis with risk scoring
• AI-powered document forensics with issue detection
• Multi-agent AI analysis system:
  • Regulatory Watcher: Monitors compliance with FINMA regulations
  • Transaction Analyst: Analyzes patterns and anomalies
  • Document Forensics: Detects digital tampering and inconsistencies
• Historical transaction context
• Complete audit trail

🏗️ Architecture

julius-baer-aml/
├── frontend/          # Next.js 14 + TypeScript + TailwindCSS
│   ├── app/          # Pages and layouts
│   ├── components/   # Reusable UI components
│   ├── lib/          # Utilities and API client
│   └── types/        # TypeScript definitions
│
├── backend/          # FastAPI + Python
│   ├── api/         # REST API endpoints
│   ├── models/      # Pydantic schemas
│   ├── services/    # Business logic
│   └── agents/      # AI agents (future)
│
└── README.md        # This file

🚀 Quick Start

Prerequisites

• Node.js 18+ (for frontend)
• Python 3.9+ (for backend)
• npm or yarn (for frontend packages)
• pip (for Python packages)

Option 1: Run Frontend Only (Recommended to Start)

The frontend works standalone with mock data - perfect for UI testing!

# Navigate to frontend
cd frontend

# Install dependencies
npm install

# Start development server
npm run dev

# Open browser to http://localhost:3000

Option 2: Run Backend Only

The backend provides REST APIs with mock data - no database required!

# Navigate to backend
cd backend

# Create virtual environment
python -m venv venv

# Activate virtual environment
# Windows:
venv\Scripts\activate
# Mac/Linux:
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

# Start server
python main.py

# API available at http://localhost:8000
# Docs at http://localhost:8000/docs

Option 3: Run Full Stack

Run both frontend and backend together:

# Terminal 1 - Backend
cd backend
python -m venv venv
source venv/bin/activate  # or venv\Scripts\activate on Windows
pip install -r requirements.txt
python main.py

# Terminal 2 - Frontend
cd frontend
npm install
npm run dev

Then open http://localhost:3000 in your browser!

📱 Screenshots & Pages

Main Dashboard (/)

• Overview of all active AML alerts
• KPI cards showing critical metrics
• Interactive charts for risk visualization
• Alert triage queue with investigation actions

Investigation Cockpit (/investigation/[alertId])

• Detailed transaction information
• Document viewer with forensics analysis
• AI agent findings from multiple specialized agents
• Historical transaction context
• Remediation and audit trail actions

🛠️ Technology Stack

Frontend

• Next.js 14 - React framework with App Router
• TypeScript - Type-safe development
• TailwindCSS - Utility-first styling
• shadcn/ui - High-quality React components
• Recharts - Data visualization
• React Query - Data fetching and caching

Backend

• FastAPI - High-performance Python API framework
• Pydantic - Data validation
• Uvicorn - ASGI server
• Mock Data - In-memory data (MongoDB ready)
• Simulated AI - Mock agents (Groq API ready)

📚 Documentation

• Frontend README - Detailed frontend documentation
• Backend README - Detailed backend documentation
• API Docs: http://localhost:8000/docs (when backend is running)

🔧 Current Status

✅ Completed

• Frontend UI with all pages and components
• Backend API with all endpoints
• Mock data for development and testing
• Complete documentation
• CORS configuration for integration

🚧 Ready to Add (When Needed)

• MongoDB integration (currently using mock data)
• Groq API integration (currently using simulated AI)
• WebSocket for real-time updates
• User authentication and authorization
• Production deployment configuration

🎨 Design Highlights

• Professional banking aesthetic matching Julius Baer's brand
• Responsive design for desktop and tablet
• Intuitive navigation between dashboard and investigation views
• Color-coded risk levels and priorities
• Interactive charts and data visualizations

🔐 Security Notes

This is a development version with:

• Mock data (no real customer information)
• No authentication (add before production)
• CORS enabled for localhost (restrict in production)

📈 Next Steps

Phase 1: Database Integration

# Install MongoDB locally or use MongoDB Atlas
# Update backend/.env with connection string
# Uncomment MongoDB code in backend/services/database.py

Phase 2: AI Integration

# Get Groq API key from https://groq.com
# Add to backend/.env: GROQ_API_KEY=your_key
# Implement real agents in backend/agents/

Phase 3: Real-Time Features

# Add WebSocket support
# Implement live alert notifications
# Add real-time dashboard updates

Phase 4: Production Deployment

# Add authentication (JWT, OAuth)
# Configure production environment variables
# Set up monitoring and logging
# Deploy to cloud (AWS, Azure, GCP)

🧪 Testing

Test Frontend

cd frontend
npm run dev
# Visit http://localhost:3000
# Click through dashboard and investigation pages

Test Backend API

cd backend
python main.py
# Visit http://localhost:8000/docs
# Try the interactive API documentation

Test Integration

# Start both frontend and backend
# Frontend will automatically connect to backend
# Test full flow: Dashboard → Click "Investigate" → View details

📝 Environment Variables

Frontend (.env.local)

NEXT_PUBLIC_API_URL=http://localhost:8000
NEXT_PUBLIC_WS_URL=ws://localhost:8000

Backend (.env)

CORS_ORIGINS=http://localhost:3000
# Future additions:
# MONGODB_URL=mongodb://localhost:27017
# GROQ_API_KEY=your_key_here

🤝 Contributing

This is a proprietary project for Julius Baer. For questions or support, contact the development team.

📄 License

Proprietary - Julius Baer

---

Built with ❤️ for Julius Baer's AML Compliance Team