SwasthAI-Guardian

Hospital Admin Dashboard Page
Outbreak Detection Page
Sakhi RAG based Women Healthcare AI Assistant
NGO Dashboard Page
Live Pad Request Page for Women (Villager -> NGO)
Disease prediction Page
Villager Dashboard Page after Login
Live Ambulance Request Page (Villager -> NGO/Admin)
Login Page
Intro Services Page

Inspiration

Over 65% of India’s population lives in rural areas where access to healthcare is still limited. We were inspired by ASHA workers who serve villages with limited medical tools, poor connectivity, and manual systems.

Our goal was to build an AI-powered healthcare platform that works even in low or zero internet conditions while supporting local languages, voice interaction, and real-time healthcare assistance.

What it does

SwasthAI Guardian is an offline-first, AI-powered healthcare platform connecting Villagers, ASHA/NGO workers, and Hospital Administrators into one unified ecosystem.

👨‍🌾 Villager Features

AI Symptom Checker with multilingual voice input and 96.8% diagnostic accuracy
Sakhi Women's Health AI powered by Grounded RAG using WHO/MoHFW guidelines
Skin Disease Scanner with image-based AI analysis
Emergency Ambulance System with GPS and offline fallback support
Sanitary Pad Request System for private NGO/ASHA assistance
Voice Input & Voice Output support
Offline PWA Support with Install-to-Home-Screen functionality
Multilingual Support for Hindi, Marathi, Tamil, Telugu, Bengali, and English

🏥 NGO / ASHA Worker Features

Maternal Health Tracker for high-risk pregnancy monitoring
Child Nutrition Monitor with BMI and malnutrition analysis
Village Health Dashboard for local healthcare statistics
AI Outbreak Alerts for disease cluster detection in assigned villages
Emergency Ambulance Feed for local emergency requests
Smart Share QR System to instantly distribute the app in rural areas
Offline Login & Sync for low-connectivity environments

🏛️ Admin Features

District Analytics Dashboard with real-time village health insights
Agentic Outbreak Radar for autonomous epidemic detection
Village Registry Management
CSV Export System for healthcare reporting
Emergency Request Monitoring
DISHA 2023 Compliant Data Management
Role-Based Access Control & Secure Reporting

How we built it

Frontend

React 18 + Vite
Tailwind CSS
Service Workers + IndexedDB
Progressive Web App (PWA)

Backend

Node.js + Express
SQLite Database
JWT Authentication

AI Services

Python FastAPI Microservice
SymptomNet Transformer Model
Groq-powered Llama 3.1
Grounded RAG Architecture
Autonomous Outbreak Detection Agent

Challenges we ran into

Building a fully offline authentication workflow
Maintaining secure background data synchronization
Optimizing AI inference for low-end mobile devices
Implementing multilingual voice interaction
Maintaining healthcare privacy compliance with DISHA standards

Accomplishments that we're proud of

Built a healthcare AI engine with 96.8% diagnostic accuracy
Developed complete offline Login, Registration, and AI diagnosis
Created an autonomous AI outbreak detection system
Added support for 6 Indian languages with voice interaction
Designed a production-grade offline-first healthcare PWA

What we learned

We learned that accessibility is as important as technology. Supporting local languages, offline functionality, and voice interaction can make healthcare technology truly usable for underserved communities.

We also learned how proactive AI systems can help improve rural healthcare response and epidemic prevention.