Razinger-Joe/epi-predict-kenya-ai

EpiPredict Kenya AI is a disease surveillance platform designed for Kenyan healthcare organizations. It helps hospitals, pharmacies, and county health departments predict disease outbreaks 14-21 days in advance with high accuracy, enabling proactive public health responses. # EpiPredict Kenya AI 🇰🇪 **AI-Powered Disease Outbreak Prediction for Kenya** Demo • Features • Architecture • Quick Start • Kubernetes • Contributing --- ## Why This Matters **The Problem:** Healthcare response to disease outbreaks in Kenya is often **reactionary**. Hospitals and counties deal with outbreaks *after* they happen, leading to overwhelmed facilities, medication shortages, and preventable loss of life. **The Pain Points:** - 📉 **Delayed Data**: Paper records take weeks to aggregate. - 🏥 **Overwhelmed Hospitals**: Sudden patient surges catch facilities off guard. - 💊 **Supply Chain Gaps**: Pharmacies run out of critical meds during peak demand. **The Solution:** EpiPredict Kenya AI flips the script from *reaction* to *prediction*. By analyzing patterns in real-time, we give decision-makers a **2-week head start** to mobilize resources, stock medicines, and warn communities. --- ## Features | Feature | Description | |---------|-------------| | 📊 **Real-time Dashboard** | Monitor disease trends across all 47 Kenyan counties | | 🔮 **Predictive Analytics** | AI-powered outbreak predictions 14-21 days ahead | | 🚨 **Smart Alerts** | Receive notifications when risk levels change | | 🧠 **ML Prediction Engine** | Gaussian Naive Bayes classifier trained on epidemiological data | | 🤖 **AI Chatbot (EpiBot)** | Context-aware LLM-powered health advisor | | 🗺️ **County Mapping** | Visualize outbreak data geographically | | 🐳 **Containerized** | Full Docker Compose & Kubernetes orchestration | | 🌙 **Dark Mode** | Full dark mode support | --- ## 🏗 Architecture The application follows a **microservices architecture** with 4 independently deployable tiers: ```mermaid graph TD Browser["🌐 Browser"] -->|Port 80 / 30080| Frontend["📱 Frontend React + Nginx"] Frontend -->|HTTP /api| Backend["⚙️ Backend FastAPI Gateway"] Backend -->|SQLAlchemy| Database[("🐘 Database PostgreSQL 15")] Backend -->|HTTP /predict, /train| MLService["🧠 ML Service FastAPI + sc …