MediKiosk 🏥

AI-Powered Smart Hospital Kiosk — Built for Smart India Hackathon (SIH) 2026

🌐 Live Demo

	Link
🖥️ Frontend (Vercel)	medi-kiosk-eight.vercel.app
⚙️ Backend API (Render)	medikiosk-afm2.onrender.com
📖 API Docs	medikiosk-afm2.onrender.com/docs

Note

The backend runs on Render's free tier — the first request after 15 minutes of inactivity may take ~30 seconds to wake up.

---

🏆 Smart India Hackathon 2026

MediKiosk was built as a submission for Smart India Hackathon (SIH) 2026, India's largest national-level hackathon organised by the Ministry of Education's Innovation Cell. The project addresses the problem of long waiting times, language barriers, and inefficient clinical intake in public hospitals across India.

Problem Statement Category: Healthcare & Biomedical Devices
Theme: Transforming Patient Experience in Government Hospitals using AI

---

🎯 What Is MediKiosk?

MediKiosk is a multilingual, AI-powered hospital kiosk platform that digitises the clinical intake process before a patient meets a doctor. It combines:

• 🤖 AI Clinical Interview — A structured 15-question health survey in 10 Indian languages
• 🚑 Ambulance Management — Real-time OSRM road-route ambulance dispatch with live tracking
• 👨‍⚕️ Doctor Dashboard — Live patient queue, AI-generated clinical summaries, and case management
• 📍 Nearby Care — Find hospitals and route to them using real road routing
• 📋 Clinical History — Automatic summarisation of patient answers into structured medical records

---

🌐 Supported Languages

Code	Language
hi	हिन्दी (Hindi)
en	English
mr	मराठी (Marathi)
bn	বাংলা (Bengali)
ta	தமிழ் (Tamil)
te	తెలుగు (Telugu)
gu	ગુજરાતી (Gujarati)
kn	ಕನ್ನಡ (Kannada)
ml	മലയാളം (Malayalam)
pa	ਪੰਜਾਬੀ (Punjabi)

All 15 clinical intake questions translate instantly offline using a curated medical dictionary. Live translation via Bhashini API when configured.

---

🏗️ Architecture

medikiosk/
├── frontend/          # React 18 + TypeScript + TailwindCSS + Vite
│   └── src/
│       ├── pages/     # Patient, Doctor, Driver, Admin portals
│       ├── components/
│       └── api/
├── backend/           # FastAPI + SQLAlchemy + SQLite
│   ├── api/           # REST endpoints
│   ├── services/      # AI engine, translation, clinical summary
│   └── models.py
└── docker-compose.yml

---

✨ Key Features

🩺 AI Clinical Intake

• Strict 15-question interview in exact medical sequence (Chief Complaint → Confirmation)
• Progress slider locked to question_index / 15 — never drifts from message count or timers
• Follow-up questions (chest radiation, medication details) don't advance the main progress
• Answers feed into structured ClinicalHistory visible to the doctor
• Red-flag detection triggers emergency alerts automatically

🚑 Ambulance System

• Real-time OSRM road-route navigation (same engine as Nearby Care)
• Live ambulance GPS simulation on a Leaflet map
• Configurable hospital destinations with government priority routing
• Full driver-side accept/reject and status flow

👨‍⚕️ Doctor Portal

• Live queue with AI-generated clinical summaries
• Verify / edit AI summaries before consultation
• Emergency alert acknowledgement
• Patient timeline and medical documents

🔐 Role-Based Access Control

• PATIENT · DOCTOR · STAFF · DRIVER · ADMIN
• JWT authentication with facility-scoped RBAC

---

🚀 Quick Start

Prerequisites

• Node.js 18+
• Python 3.11+
• (Optional) Docker & Docker Compose

1. Clone

git clone https://github.com/divyamc1803/MediKiosk.git
cd MediKiosk

2. Backend

cd backend
python -m venv venv && source venv/bin/activate
pip install -r requirements.txt
python seed_demo.py        # seeds demo accounts
uvicorn main:app --reload --port 8000

3. Frontend

cd frontend
npm install
npm run dev                # → http://localhost:5173

4. Docker (one-command)

docker-compose up --build

---

🎭 Demo Accounts

Role	Email	Password
Patient	aarav@demo.com	demo123
Doctor	sneha@demo.com	demo123
Driver	driver@demo.com	demo123
Admin	admin@demo.com	demo123

---

🔑 Environment Variables (Optional)

Variable	Description
BHASHINI_API_KEY	Bhashini live translation API key
BHASHINI_TRANSLATION_SERVICE_ID	Bhashini service ID
SECRET_KEY	JWT signing secret

The app runs fully offline without Bhashini credentials using a built-in curated clinical translation dictionary.

---

🧪 Running Tests

# Backend API verification
cd backend
python scratch/verify_intake_direct.py

# Frontend type-check
cd frontend
npx tsc --noEmit

---

📸 Screenshots

Clinical Intake — Question 6 of 15 in Hindi, progress 40% Doctor Dashboard — AI clinical summary ready for review Ambulance Demo — Real-time OSRM road routing

---

🤝 Team

Built with ❤️ for Smart India Hackathon 2026 by Team MediKiosk.

Member	GitHub
Divyam Choudhary	@divyamc1803
Aditya Bathla	@Aditya-bathla

---

📄 License

MIT © 2026 Team MediKiosk