Vocal AI-powered dementia assessment system with conversational interface with Simli avatars and MMSE-based severity classification.
🌐 Live Demo: https://caresmafe-production.up.railway.app/
- Interactive Voice Conversation: Real-time voice interaction with AI avatar
- Automated Assessment: MMSE-based cognitive scoring across 5 domains
- Visual Avatar: Simli-powered avatar with lip-sync
- Care Plan Generation: Personalized recommendations based on severity
- CLI Tool: Standalone script for analyzing existing transcripts
# Backend
cd backend
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
# Frontend
cd ../frontend
npm install# Backend .env
cd backend
cp .env.example .env
# Edit .env and add: OPENAI_API_KEY, INWORLD_API_KEY, CARTESIA_API_KEY
# Frontend .env
cd ../frontend
cp .env.example .env
# Edit .env and add: VITE_SIMLI_API_KEY, VITE_SIMLI_FACE_ID (optional)# Terminal 1: Start backend
cd backend
source venv/bin/activate
uvicorn app.main:app --reload
# Terminal 2: Start frontend
cd frontend
npm run devVisit http://localhost:5173 to use the application.
Analyze existing transcripts without running the full application:
cd backend
source venv/bin/activate
# Analyze a transcript
python3 cli.py --path <fileName>.txt
python3 cli.py --path <fileName>.json
# Output as JSON
python3 cli.py --path <fileName>.json --output-jsonSupports .txt and .json input files.
| Score | Severity | Follow-up |
|---|---|---|
| 25-30 | Normal | 12 months |
| 20-24 | Early | 6 months |
| 10-19 | Mid | 1 month |
| 0-9 | Late | 1 week |
cd backend
source venv/bin/activate
# Run all tests
pytest
# Run without API calls
pytest -m "not integration"- OpenAI: platform.openai.com
- Inworld: inworld.ai
- Cartesia: cartesia.ai
- Simli: simli.com
This is an automated screening tool, not a clinical diagnosis. Professional evaluation by qualified healthcare providers is required for any medical decisions.