MedVerse

AI-powered healthcare platform connecting Patients, Doctors, Laboratories, and Admin — built with Flutter, Flask, Firebase, MongoDB GridFS, and Google Gemini AI.

---

Architecture

[ Flutter App ] ──► [ Flask API ] ──► Firebase Auth (Token Verify
                                  ──► MongoDB GridFS (File Storage)
                                  ──► Firestore (Profiles & Appointments)
                                  ──► Google Gemini AI (Document Analysis)

---

Key Features

Multi-Role Platform

• Patient — Book appointments, upload documents, view reports, AI chatbot
• Doctor — Manage appointments, review patient records & lab reports
• Laboratory — Upload lab reports for patients
• Admin — Verify doctor/lab registrations, platform management

Smart Booking System

• Doctor appointment & lab test booking with time-slot selection
• Duplicate request blocker (1 active request per doctor/lab per day)
• Status flow: Pending → Approved → Checked-In → Completed
• Real-time sync across patient & provider dashboards

AI-Powered Features

• MediDoc Analyze — Classifies & extracts data from prescriptions, lab reports, bills, discharge summaries
• AI Health Chatbot — Symptom queries, medicine info, nutrition guidance (Gemini 2.5 Flash)
• Persistent chat sessions with full conversation history

Secure Document Storage

• All files are stored in MongoDB GridFS (not Firestore)
• Token-based API auth (Bearer <Firebase ID Token>)
• Doctors can only access patient files if an active appointment exists
• Patients can upload, view, download & delete their own documents

Premium UI/UX

• Glassmorphism design with light/dark theme support
• Interactive card-first doctor dashboard
• Inline loading spinners & anti-double-click guards
• Collapsible medical history timeline
• Real-time push notifications (FCM)

Auth & Security

• Firebase Authentication with mandatory email verification
• Role-based access control for privacy
• Lazy Firestore document creation (post-verification only)
• Secure password reset flow

---

🛠️ Tech Stack

Layer	Technology
Frontend	Flutter (Dart), Google Fonts, Syncfusion PDF Viewer
Backend	Python, Flask
AI	Google Gemini 2.5 Flash
Auth	Firebase Authentication
Database	MongoDB Atlas (GridFS), Cloud Firestore
Notifications	Firebase Cloud Messaging (FCM)
Hosting	Render

---

Project Structure

MedVerse/
├── backend/
│   ├── app.py                    # Main Flask server
│   ├── ai_engine.py              # Gemini AI document analyzer
│   ├── firebase_config.py        # Firebase Admin SDK config
│   ├── db.py                     # MongoDB connection
│   ├── routes/
│   │   ├── __init__.py           # Core API routes
│   │   └── storage_routes.py     # GridFS storage endpoints
│   └── services/
│       └── storage_service.py    # GridFS CRUD operations
├── frontend/
│   └── lib/
│       ├── main.dart             # App entry & routing
│       ├── screens/              # Patient, Doctor, Lab, Admin UIs
│       ├── services/             # API & Firebase services
│       ├── widgets/              # Reusable glass components
│       └── theme/                # Glassmorphism & app themes
└── README.md

---

🔌 API Endpoints

All routes require Authorization: Bearer <ID_TOKEN> header.

Method	Endpoint	Description
POST	/api/storage/upload	Upload file (multipart)
GET	/api/storage/patient	Get own documents
GET	/api/storage/patient/<id>/documents	Doctor: view patient docs
GET	/api/storage/patient/<id>/lab-reports	Doctor: view lab reports
GET	/api/storage/file/<fileId>	Download file
DELETE	/api/storage/delete/<fileId>	Delete file

---

⚙️ Setup

Backend

cd backend
python -m venv venv
venv\Scripts\activate          # Linux/Mac: source venv/bin/activate
pip install -r requirements.txt

Create backend/.env:

PORT=5000
MONGO_URI=mongodb+srv://<user>:<pass>@cluster.mongodb.net/
DB_NAME=medverse
FIREBASE_CREDENTIALS_PATH=your-firebase-adminsdk.json
GEMINI_API_KEY=your_gemini_key

python app.py

Frontend

cd frontend
flutter pub get

Update lib/config/api_config.dart with your backend URL, then:

flutter run

Production (Render)

• Store Firebase credentials as Base64 in FIREBASE_CREDENTIALS_JSON env var
• Whitelist 0.0.0.0/0 in MongoDB Atlas Network Access