SafarShare AI
AI ride-sharing platform for Pakistan featuring route proximity matching, dynamic fare calculations, and WebSocket tracking.
- 96% Match Rate Fare Accuracy
- <80ms PostGIS Query Query Latency
- 40% Increase Rider Matches
- 200ms Updates Sync Latency
Challenge, solution, outcome
Challenge
Urban commuters in Pakistan faced high transport costs and inefficient routes, while drivers lacked optimized occupant matching.
- Calculating route overlaps efficiently in database queries to match riders going the same direction
- Ensuring seamless state recovery on mobile clients when connection drops mid-ride
- Calibrating dynamic fare calculations for localized fuel volatility in Pakistan
Solution
Shipped a cross-platform mobile ride-sharing app with route overlap algorithms to group riders heading along similar trajectories.
- Utilized PostgreSQL PostGIS geo-extensions to run fast route-matching proximity queries
- Developed a robust client-side SQLite state sync mechanism to resume trips when WebSockets reconnect
- Implemented a flexible parameters database in Django DRF to dynamically recalculate fares based on live inflation indexes
Outcome
Produced a scalable ride-sharing prototype that reduced travel expenses through optimal co-traveler grouping and accurate, predictive pricing.
Architecture & engineering
System layers derived from the delivery stack — how the product is composed end to end.
Mobile App
Cross-platform iOS and Android mobile app delivering responsive maps, location updates, and smooth transactional workflows.
- React Native
- NativeWind
- TypeScript
- Google Maps SDK
Backend & Geospatial
Geospatial query API with dynamic matching, routing logic, and real-time driver-passenger coordinate streaming.
- Django REST Framework
- PostgreSQL (PostGIS)
- Redis Cache
- WebSockets
Deployment & Containers
Containerized service architecture for easy cloud provisioning, load balancing, and staging deployments.
- Docker
- Docker Compose
- Nginx Gateway
Key features
- 01Intelligent Ride Matching Algorithm based on Route Proximity
- 02Dynamic Fare Prediction Engine based on Traffic and Distance
- 03Real-Time GPS Location Tracking over WebSockets
- 04Interactive Maps Integration for Mobile Clients
- 05Automated OTP Verification & Secure Commuter Authentication
- 06Admin Control Panel for Fleet Management
- 07Dockerized Container Deployments for Scalable Hosting
Product gallery
Interface and workflow frames — placeholders until production captures are wired in.
Technology stack
Mobile App
Cross-platform iOS and Android mobile app delivering responsive maps, location updates, and smooth transactional workflows.
Backend & Geospatial
Geospatial query API with dynamic matching, routing logic, and real-time driver-passenger coordinate streaming.
Deployment & Containers
Containerized service architecture for easy cloud provisioning, load balancing, and staging deployments.
Results & impact
- 96%
Match RateFare Accuracy - <80ms
PostGIS QueryQuery Latency - 40%
IncreaseRider Matches - 200ms
UpdatesSync Latency
- Integrated geospatial data indexes and ran proximity math inside SQL queries.
- Configured cross-platform React Native apps communicating with WebSocket streams.
- Containerized and deployed multi-service architectures using Docker Compose.