margaretnduta/Weatherwise_Ai

An AI-powered weather decision-support platform that turns raw weather data into practical, crop-specific farming advice for smallholder farmers in Kenya. # WeatherWise AI An AI-powered weather decision-support platform that turns raw weather data into practical, crop-specific farming advice for smallholder farmers in Kenya. Built for **Hack the Weather 2026 Hackathon**. --- ## Problem Farmers have access to weather forecasts, but not actionable, localized recommendations. Knowing it will rain 20% of the time doesn't tell a farmer in Trans Nzoia whether to plant, irrigate, or wait. **WeatherWise AI** closes that gap by converting weather data into direct farming actions. ## Solution WeatherWise AI combines: - **Rainfall Prediction** — ML model (Random Forest/XGBoost) forecasting rainfall probability and intensity - **Recommendation Engine** — rule-based logic converting forecasts into farming actions (plant, irrigate, fertilize, spray, harvest) - **AI Weather Copilot** — LLM + RAG assistant answering natural-language farming questions - **Smart Alerts** — real-time notifications for drought, heavy rainfall, wind, and heat stress - **Interactive Dashboard** — weather trends, crop risk levels, and county-level risk maps ## Core Features | Feature | Description | |---|---| | 🌤 Weather Summary | Real-time temperature, rainfall, humidity, wind, UV index | | 📊 Weather Trends | Historical + forecasted rainfall charts | | 🌽 Crop Health & Drought Risk | Predicted stress/drought risk levels | | 🤖 AI Recommendations | Personalized planting/irrigation/fertilizing advice | | 🚨 Smart Alerts | Notifications for extreme weather events | | 🗺️ Interactive Weather Map | County-level risk visualization | | 💬 AI Weather Copilot | Natural-language Q&A on farming decisions | ## Tech Stack - **Language:** Python - **ML:** scikit-learn, XGBoost - **AI Copilot:** LLM + RAG (LangChain, ChromaDB) - **Visualization:** Plotly, Folium - **App/UI:** Streamlit - **Prototyping:** Jupyter Notebooks ## Project Structure ``` weatherwise-ai/ ├── data/ # raw, processed, and sample weather data ├── notebooks/ # expl …