Skip to content

README-RISK-MODEL.md

Latest commit

 

History

History
78 lines (62 loc) · 3.8 KB

README-RISK-MODEL.md

File metadata and controls

78 lines (62 loc) · 3.8 KB
Raw

Risk Model Management System - MongoDB Integration

Overview

This project implements a Risk Model Management system for fraud detection using MongoDB as the database. The implementation showcases key advantages of MongoDB over traditional SQL databases, particularly:

  1. Schema Flexibility: Add custom risk factors without schema migrations
  2. Real-Time Updates: Use Change Streams for instant notifications
  3. Document Model: Natural data representation with embedded risk factors
  4. Versioning: Complete model versions stored as separate documents

Implementation Files

Backend

  • /backend/routes/model_management.py: API endpoints for model CRUD operations
  • /backend/services/risk_model_service.py: Service for managing risk models and Change Streams
  • /backend/dependencies.py: MongoDB connection and service initialization

Frontend

  • /frontend/components/ModelAdminPanel.js: UI component for model management with real-time updates

Features

MongoDB-Powered Features

  • Dynamic Risk Factors: Add new risk factors on-the-fly without schema migrations
  • Real-Time Model Updates: WebSocket + Change Streams for instant notifications
  • Complete Document Model: All model data in a single document without JOINs
  • Versioned Models: Automatic versioning with full model history using MongoDB sort capabilities to ensure proper version sequencing
  • Performance Tracking: Risk factor distribution and effectiveness metrics

API Endpoints

  • GET /models/: List all risk models
  • GET /models/{model_id}: Get a specific risk model
  • POST /models/: Create a new risk model
  • PUT /models/{model_id}: Update an existing model
  • DELETE /models/{model_id}: Archive a risk model
  • POST /models/{model_id}/activate: Activate a specific model
  • GET /models/{model_id}/performance: Get performance metrics
  • POST /models/{model_id}/feedback: Record transaction outcome feedback
  • WebSocket /models/change-stream: Real-time model updates via Change Streams

Demo Flow

For detailed demo instructions, see RISK_MODEL_MANAGEMENT_IMPLEMENTATION.md

Quick Demo Steps

  1. Model Selection: View and select different risk models
  2. Schema Flexibility: Add a custom risk factor without schema migrations
  3. Real-Time Updates: See changes instantly across multiple sessions
  4. Document Model: Examine the complete model structure
  5. Version Management: Create a new version and activate models
  6. Performance Metrics: View risk factor distribution and effectiveness

MongoDB Advantages Highlighted

  • Add new fields without migrations vs. SQL ALTER TABLE operations
  • Real-time updates via Change Streams vs. SQL polling or triggers
  • Natural document structure vs. SQL table joins
  • Simple versioning with complete documents vs. SQL change tracking

Setup Requirements

  • MongoDB Atlas cluster (M0 or higher)
  • Python 3.9+ with FastAPI and Motor/PyMongo
  • Node.js with React for the frontend
  • WebSockets enabled (install websockets library if needed)

Running the Application

  1. Install dependencies: poetry install (backend) and npm install (frontend)
  2. Configure MongoDB connection in .env
  3. Start backend: poetry run uvicorn main:app --reload
  4. Start frontend: npm run dev
  5. Navigate to Risk Model Management page in the application

For detailed implementation information, see RISK_MODEL_MANAGEMENT_IMPLEMENTATION.md

Related Documentation