Jade-Isimbi/GUIDAPLATE

# GuidaPlate AI-Powered Dietary Decision Support for CKD Patients in Rwanda BSc Software Engineering Capstone African Leadership University · July 2026 ## Links | | | |---|---| | 🎥 Demo Video | drive.google.com | | 🌐 Live App | guidaplate.vercel.app | | 🔌 Backend API | guidaplate-production.up.railway.app — **redeploy may be required** (see Live URLs) | | 📁 Repository | github.com | --- ## Table of contents 1. System requirements 2. Moderator evaluation flow 3. Quick start (local) 4. Demo account 5. Prerequisites 6. Model file availability 7. How to run — detailed 8. Environment variables 9. Verify the stack is healthy 10. Running tests 11. Production deployment 12. Key results 13. Known limitations 14. Project structure 15. Architecture 16. Notebooks 17. Features 18. Troubleshooting 19. License --- ## System requirements Before running GuidaPlate, ensure you have: - **Python 3.11** - **Node.js 18+** - **npm** - **Git** - **At least 5 GB free storage** for ML dependencies (TensorFlow / XGBoost wheels, `node_modules`, model files) The project was tested on **macOS (CPU-only)** and **cloud deployment** environments (Vercel frontend + Railway backend). No GPU is required. --- ## Moderator evaluation flow A reviewer can evaluate the project using the following sequence: 1. **Clone** the repository 2. Confirm **model files** are present (`git ls-files models/` — see Model file availability) 3. Copy `.env.example` → `.env` and set at least `JWT_SECRET`, `ALLOWED_ORIGINS`, and (for chat) `GROQ_API_KEY` 4. **Install backend** dependencies (`pip install -r requirements.txt` in a Python 3.11 venv) 5. **Install frontend** dependencies (`cd frontend && npm install`) 6. **Start backend API** (`uvicorn backend.main:app --reload --port 8000`) 7. **Start frontend** (`npm run dev`) and open localhost 8. **Create an account** on the registration page (accept …