NAB Payment Safety Assistant

A Flask-based AI chatbot that helps users identify potentially risky payment requests and scam transactions through conversational AI and expert risk assessment.

Features

• Interactive Scam Detection: 4-question assessment covering payment recipient, purpose, source, and platform verification
• AI-Powered Risk Analysis: OpenAI GPT-powered scam risk assessment with LOW/MEDIUM/HIGH classification
• NAB Messaging Interface: Professional banking UI matching NAB brand standards
• STOP-CHECK-PROTECT Framework: Standardized safety recommendations following established fraud prevention guidelines
• Real-time Chat Interface: Mobile-responsive design with avatar-based messaging
• Automatic Banker Follow-up: Integration with NAB customer service workflow

Tech Stack

• Backend: Python Flask with OpenAI GPT-3.5-turbo integration
• Frontend: Responsive HTML/CSS/JavaScript with NAB branding
• Deployment: Render cloud platform with GitHub CI/CD
• AI Model: OpenAI GPT-3.5-turbo for intelligent scam detection

Quick Start

Local Development

1. Clone the repository:

git clone https://github.com/xichaow/fraud-chatbot.git
cd fraud-chatbot

2. Install dependencies:

pip install flask openai

3. Set up environment variables:

export OPENAI_API_KEY="your-openai-api-key-here"

4. Run the application:

python app.py

5. Access the application:

• Homepage: http://localhost:8000/
• Chat Interface: http://localhost:8000/chat

Production Deployment

The application is configured for Render deployment with automatic GitHub integration.

Required Environment Variables:

• OPENAI_API_KEY - Your OpenAI API key (required)
• PORT - Port number (defaults to 8000)
• DEBUG - Debug mode (defaults to False)

How It Works

User Journey

1. Welcome: Professional NAB-branded chat interface
2. Assessment: 4 targeted questions about payment details:
   • Who are you paying?
   • What is the payment for?
   • Where did you get the payment link?
   • Which website/platform are you using?
3. Analysis: AI-powered scam risk assessment
4. Recommendations: STOP-CHECK-PROTECT safety guidelines
5. Follow-up: Automatic NAB banker callback notification

Risk Assessment Framework

Risk Levels:

• LOW: Payment appears legitimate with minimal risk indicators
• MEDIUM: Some concerning factors requiring additional verification
• HIGH: Multiple red flags indicating likely scam attempt

Assessment Criteria:

• Recipient legitimacy and verification
• Payment purpose and context analysis
• Source credibility of payment requests
• Platform/website authenticity checks

Safety Recommendations

STOP-CHECK-PROTECT Framework:

• STOP: Don't give money or information to anyone if unsure. Scammers pretend to be from trusted organizations.
• CHECK: Ask yourself if the message or call is fake. Never click links in messages. Only contact businesses using official website or app contact information.
• PROTECT: Act quickly if something feels wrong. Contact your bank immediately if you notice unusual activity or if a scammer gets your information.

API Documentation

Endpoints

• GET / - Landing page with service overview and scam education
• GET /chat - NAB Messaging chat interface
• POST /api/chat/start - Initialize conversation session
• POST /api/chat/respond - Process user responses and provide next questions
• GET /static/<filename> - Serve static assets (NAB branding)
• GET /health - Application health check

Response Format

Risk Assessment Output:

**RISK LEVEL: [LOW/MEDIUM/HIGH]**

• [Key risk factor or positive indicator 1]
• [Key risk factor or positive indicator 2] 
• [Key risk factor or positive indicator 3]
• Recommendation: [STOP-CHECK-PROTECT framework guidelines]

Features & Enhancements

Current Features

✅ NAB Messaging UI with professional branding
✅ Real-time chat with avatars and timestamps
✅ Intelligent response validation
✅ Structured risk assessment output
✅ STOP-CHECK-PROTECT safety framework
✅ Automatic banker follow-up workflow

Future Enhancements

🔄 NAB Data Integration: Plans to incorporate NAB's internal fraud database and transaction patterns for enhanced accuracy
🔄 Advanced Analytics: Enhanced scam pattern recognition
🔄 Multi-language Support: Expand accessibility

Project Structure

fraud-chatbot/
├── app.py                    # Main Flask application
├── static/
│   └── nab-icon.jpg         # NAB branding assets
├── example/
│   ├── chatbot_snapshot.jpg  # UI design references
│   ├── chatbot_snapshot2.png
│   ├── chatbot_snapshot3.PNG
│   ├── customer_chatbot_snapshot.png
│   ├── icon.jpg
│   └── recommendation.jpg   # STOP-CHECK-PROTECT framework
├── requirements.txt         # Python dependencies
├── render.yaml             # Render deployment configuration
└── README.md               # Project documentation

The 4 Scam Detection Questions

1. Payment Recipient: "Who are you making this payment to? Please provide the name of the person, organization, or company."

2. Purpose of Payment: "What is the purpose of this payment? Please describe what you are paying for (service, product, investment, etc.)"

3. Source of Payment Link: "Where did you get the payment link or payment instructions from? Please share the source (email, website, text message, social media post, etc.)"

4. Website Verification: "Please provide the website URL or platform where you are making this payment, or describe how you are accessing the payment page."

Testing

Local Testing

# Run the application locally
python app.py

# Test the chat interface
open http://localhost:8000/chat

# Test API endpoints
curl -X POST http://localhost:8000/api/chat/start \
  -H "Content-Type: application/json" \
  -d '{"session_id": "test123"}'

Health Check

curl http://localhost:8000/health

Contributing

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

Security & Privacy

• No persistent storage of user conversations
• Session-based temporary data only
• Secure API key management
• Input validation and sanitization
• No logging of sensitive payment information

License

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

Support

For technical support or questions about the NAB Payment Safety Assistant, please contact the development team or create an issue in the GitHub repository.