Marcelofury/UGANDA-VHT-Co-Pilot

# Uganda VHT Co-Pilot 🇺🇬 > **AI-Enabled Community Health Tool** — Simplifying diagnostics, detecting malnutrition via computer vision, and delivering real-time outbreak alerts across Uganda's districts. --- ## 🚀 Quick Start ```bash # Install dev server npm install # Run locally (opens at localhost) npm run dev ``` ## 🔑 Setup (First Time) 1. **Get a free Groq API Key** → console.groq.com 2. Open the app → enter your key in the popup modal 3. Or click **"Use Demo Mode"** to explore without a key --- ## 📁 Project Structure ``` VHTcopilot/ ├── public/ │ ├── index.html ← Main PWA app │ ├── style.css ← Blue/white design system │ ├── app.js ← All JS logic (Groq, CV, Map, Alerts) │ ├── sw.js ← Service worker (offline) │ └── manifest.json ← PWA installable manifest ├── firebase.json ← Firebase Hosting config ├── firestore.rules ← Firestore security rules ├── .firebaserc ← Firebase project alias └── package.json ``` --- ## 🧠 AI Features | Feature | Technology | Status | |---------|-----------|--------| | Symptom Diagnostics | Groq API · LLaMA 3.3 70B | ✅ Live | | Malnutrition Vision | TF.js + Vision API (CV Pipeline) | 🔬 Prototype | | Outbreak Alerts | Firebase + Groq summaries | ✅ Demo | | District Map | Leaflet.js + OpenStreetMap | ✅ Live | | Offline Mode | Service Worker PWA | ✅ Live | --- ## 🖥️ Deploy to Firebase Hosting ```bash # Install Firebase CLI npm install -g firebase-tools # Login firebase login # Initialize (first time only) firebase init # Deploy npm run deploy # → Live at: vht-copilot-uganda.web.app ``` --- ## 💻 Computer Vision Pipeline ``` Camera/Photo → MediaPipe Body Detection → MobileNetV2 Classification → Google Vision API Validation → Groq AI Recommendation ``` **MUAC Thresholds (WHO):** - 🔴 12.5 cm → Normal — Monthly monitoring --- ## 🚨 Outbreak Alert Thresholds | Disease | Alert Trigger | |---------|--------------| | Malaria | 15+ cases / …