SWASTHYA
DOOT
Empowering ASHAs and ANMs with an offline-first resilient mobile app, backed by a massive Intelligent Layer Hub for disease forecasting and risk clustering.
The Reality of the Field
Rural Healthcare is
Disconnected & Manual
Field health workers spend over 40% of their time on manual paperwork, delaying critical care. Furthermore, deep rural areas entirely lack reliable internet, rendering traditional cloud-dependent health registries useless precisely where they are needed most.
High data entry burden causing delayed risk identification during pregnancies.
Zero connectivity in deep rural sectors breaking sync protocols.
Lack of systemic insights for government allocation & outbreak prevention.
The Twilight Architecture
An Intelligent Offline Ecosystem
Swasthya Doot is not a simple app; it is a multi-tier resilient architecture. It uses Isar local DBs for strict offline-first data capture, background workers to silently upload to Firestore, and a massive Python Intelligent Layer to cluster risk for Government Dashboards.
Offline-First Flutter companion using Isar + Workmanager background sync.
On-device Google ML Kit OCR to instantly parse Aadhaar tracking data.
Intelligent Hub running ARIMA forecasting & Neo4j Knowledge Graphs.
Built-in Gamification Engine (Streaks/Leaderboards) for worker retention.
The Platform Capabilities
Engineered for edge environments and massive scale state-level aggregation.
The "Ask Didi" Deep AI
A specialized Neo4j Knowledge Graph maps complex villager relationships (Mother→Child→Risk) to power field workers with instant diagnostic context.
Hover nodes to traverse Knowledge Graph
Resilient Offline DB
Runs Isar local database natively on the device. All CRUD operations happen instantly locally, and a dedicated Workmanager handles the Firebase sync queue silently in the background when 4G is restored.
On-Device OCR
Integrated Google ML Kit directly onto the edge device to parse Aadhaar cards physically, radically accelerating family registry onboarding while keeping PII secure off the cloud.
Adoption Gamification
Features an internal points system and regional leaderboards. ASHAs earn rep for tracking ANC (Antenatal Care) schedules, driving massive daily active usage through competitive incentives.
Intelligent Layer Hub
Uses predictive models (ARIMA / LSTM) to forecast disease trends like diabetes/anemia across blocks, and K-Means clustering to autodetect highly vulnerable household risk pockets.
Multi-Tier Dashboards
Raw edge data is aggregated into dynamic dashboards serving the PHC (Block Tracking), the CMO (District Risk Alerts), and the NHM HQ (State Disease Forecasting & Visualizations).
Engineered to Scale State-Wide.
Swasthya Doot demonstrates our mastery over complex, multi-environment architectures. We successfully bridged the gap between zero-connectivity edge devices and massive AI-driven cloud aggregation.