Projects
Urgent Blood App
Urgent Blood App is a high-performance emergency response platform designed to optimize blood donation during critical time windows. It implements a smart matching algorithm using the Haversine formula for precise geo-location sorting, ensuring donors are alerted based on proximity and eligibility. The system integrates real-time communications via Socket.io and Firebase Cloud Messaging (FCM) to facilitate instant coordination between requesters and donors.
The problem this solves
Blood shortages during emergencies cost lives — not because blood isn't available, but because the logistics of finding compatible nearby donors fast enough consistently fail. Existing solutions are too slow and too manual for genuine emergencies. This platform is built specifically for the time-critical scenario: match the right donor, in the right location, as fast as possible.
What it does
- Haversine-based geo-matching to surface the closest compatible donors first
- Real-time push notifications via Firebase Cloud Messaging
- Live request feeds with Socket.io for instant updates without polling
- Blood group compatibility checking baked into matching logic
- Donor trust and reputation system to improve match reliability
- AI-driven priority scoring for critical vs. standard requests
- Google Maps integration for visual donor proximity display
- Auto-escalation for unfulfilled requests past a time threshold
Hard parts and how I solved them
The core challenge was latency — in an emergency, every second matters. Getting the matching, notification, and confirmation flow to complete in under 10 seconds required careful architecture decisions: real-time sockets over REST for status updates, FCM for push delivery, and pre-computed geo-indexes in MongoDB to avoid slow proximity queries under load. Building the trust system was also non-trivial — a donor who repeatedly accepts requests but doesn't show up needs to be surfaced less prominently without being permanently penalized.
Tech stack
Outcome
The platform is in active development with the core matching and notification pipeline working end-to-end. The geo-matching algorithm reliably identifies the closest eligible donors within milliseconds, and the real-time communication layer keeps all parties synchronized without manual refresh.
My role
Lead developer — system architecture, geo-matching algorithm, real-time socket layer, push notification integration, React Native mobile app, and the Node.js/Express backend.