How abdullah-naeem-gh/crime_analysis_backend works
Project Type & Purpose
• Type: Node.js Express backend API.
• Purpose: Manage, analyze, and predict crime data; support community reports; compute safest routes; and provide AI-powered crime pattern predictions.
a. API Gateway / Express Server
• Entry point: src/index.js
• Hosts all route modules (“routes/auth.js” and src/routes/{ai, analytics, safePath}).
b. Route Handlers
• Authentication (OAuth)
• CRUD APIs for crimes, criminals, victims, stations (Supabase)
• Community Reports & Comments (MongoDB)
• Analytics (/api/analytics)
• AI Predictions (/api/ai)
• Safest Path Calculation (/api)
c. Business-Logic & Helpers (utils/)
• supabaseClient.js → PostgreSQL via Supabase
• mongodbClient.js → MongoDB
• redisClient.js → Redis cache
• geminiClient.js → Google Gemini API
• analyticsDataFetcher.js → Aggregates analytics data for AI & pathfinding
• pathfindingAlgorithm.js → OpenRouteService integration + scoring
d. External Systems
• Supabase (PostgreSQL)
• MongoDB
• Redis
• Google OAuth 2.0
• Google Gemini API
• OpenRouteService API
Relationships & Data Flows
Example: /api/crimes → supabaseClient → Supabase DB.
• For “calculate-safest-path”, route → analyticsDataFetcher → pathfindingAlgorithm → OpenRouteService → returns to client.
Architectural Patterns & Principles
• Layered Separation:
• Presentation/API layer (Express routes)
• Service/Business logic (analyticsDataFetcher, pathfindingAlgorithm, AI orchestration)
• Data-access layer (supabaseClient, mongodbClient, redisClient, geminiClient)
• Stateless API with external session/auth (Google OAuth).
• Caching layer (Redis) to accelerate AI predictions.
• External API integrations for AI and mapping.