A prototype wellness tracking app that collects daily self-reports through questionnaires and multiple-choice inputs, then generates a personalized wellness score. Built to explore how off-the-shelf large language models (LLMs) can transform subjective symptom narratives into structured, clinically relevant insights. This repository serves as an open prototype for developing tools that bridge patient self-reporting and actionable wellness metrics.
- Install dependencies:
pip install -r requirements.txt- Copy environment configuration:
cp .env.example .env-
(Optional) Add your OpenAI API key to
.envfor enhanced AI insights. Without it, the app uses a local keyword-based fallback.- Copy
.env.exampleto.envand setOPENAI_API_KEY. - Optionally set
OPENAI_MODEL(default:gpt-3.5-turbo).
- Copy
-
Run the application:
python app.py- Open http://localhost:5000 in your browser
- Daily Self-Reports: Interactive questionnaires with multiple-choice inputs
- Wellness Score: Personalized scoring algorithm based on mood, energy, sleep, and stress
- AI Insights: LLM-powered analysis of symptom narratives (OpenAI Chat Completions when configured; keyword-based fallback otherwise)
- Progress Tracking: Historical data and trend analysis
- User-Friendly Interface: Responsive web interface with real-time feedback
POST /api/users- Create a new userPOST /api/users/{id}/reports- Submit daily wellness reportGET /api/users/{id}/reports- Get wellness reportsGET /api/users/{id}/wellness-summary- Get wellness summary and trends
- Backend: Flask, SQLAlchemy
- Frontend: HTML5, CSS3, JavaScript, Bootstrap
- Database: SQLite (easily configurable to PostgreSQL/MySQL)
- AI Integration: OpenAI Chat Completions (env-configurable) with safe fallback
- Advanced LLM integration for deeper symptom analysis
- Data visualization with charts and trends
- Mobile app development
- Integration with wearable devices
- Multi-user support with authentication