Samraksha is a cutting-edge real-time Police Crime Monitoring and Case Management system. It seamlessly integrates a powerful Python backend with a high-performance React front end, leveraging modern AI/ML and geospatial techniques to give law enforcement an unprecedented edge in operational capabilities.
- Real-time Map & Routing: Features dynamic, real-time routing logic that generates exact road-based paths using locally hosted OSRM.
- AI/ML Patrol Dispatching: Uses Google OR-Tools to dynamically re-calculate and re-route police patrols precisely along real-world road networks, prioritizing areas with active hotspots.
- Dynamic Database-backed Infrastructure: Every piece of data—from unit locations, live incident alerts, officer credentials, and IAM-style granular permissions (PBAC)—is live and directly fed from PostgreSQL. There are NO hardcoded mocks masking data outages.
- Analytics & Predictive Insights: Tracks historical trends and forecasts localized crime hotspots using data science analytics models (XGBoost, DBSCAN), helping commanding officers pre-deploy units automatically.
- Live Notifications & Control Center: Integrates live WebSocket events (with exponential backoff reconnects) so officers are notified immediately of changes, PCR webhooks, and routing re-assignments.
- AI Assistant: Natural language querying over case documents and data using Llama.cpp and Whisper for voice search.
- CCTNS Integration: Configured endpoints for interfacing with the national CCTNS network.
- Frontend: React 19, Vite, Leaflet, Zustand, Axios
- Backend: FastAPI, Asyncpg, SQLAlchemy, XGBoost, Scikit-learn
- Databases: PostgreSQL (with PostGIS for GeoSpatial mapping) and Redis
- Routing Engine: OSRM + Google OR-Tools
All services are containerized via Docker and orchestrated using Docker Compose.
Run docker compose up -d --build to deploy the entire stack instantly.