- Project
- 01
- Category
- AI / Healthcare
- Period
- March 2026
SehatConnect
Privacy-Aware Rural Healthcare Access System
Overview
A low-bandwidth healthcare system reaching patients without smartphones — IVR, offline reminders and voice support alongside a full mobile client.
- 01
Designed a low-bandwidth healthcare system with IVR, offline reminders, WebRTC, and Node.js/MongoDB.
- 02
Built accessibility-first workflows for non-smartphone users through IVR and voice support.
- 03
Integrated AI workflows for prescription verification, chatbot support, and hospital/pharmacy discovery.
Problem
Rural healthcare access is constrained long before software is: intermittent connectivity, feature phones, and patients who cannot rely on an app being reachable when they need it. A conventional mobile-only product excludes the people it is meant to serve.
Approach
- 01
Treat the phone call as a first-class interface, not a fallback — IVR and voice workflows carry the full core journey.
- 02
Keep state local first. Reminders and records work offline through SQLite and sync when a connection returns.
- 03
Run inference on-device with TensorFlow Lite where the network cannot be trusted.
- 04
Use WebRTC for consultations so sessions degrade gracefully rather than failing outright.
Architecture
- 01AccessTwilio IVR + React Native client
- 02Local stateSQLite — offline reminders and records
- 03ConsultationWebRTC session layer
- 04IntelligenceTensorFlow Lite on-device inference
- 05ServicesNode.js API, MongoDB persistence
Technology
- React Native
- Node.js
- MongoDB
- TensorFlow Lite
- WebRTC
- SQLite
- Twilio IVR
Project details
Details available on request.
Gallery
