Predictive Intelligence & Crop Protection via Climatic Analytics
AgriSense AI is a proactive, cloud-native web application designed to shift agricultural practices from a reactive wait-and-see approach to a preventative, data-driven methodology. By combining hyper-local 7-day weather forecasting with a Random Forest machine learning model, the platform predicts crop disease risks and provides actionable treatment protocols before damage occurs.
- Real-Time Climatic Integration: Fetches live and 7-day predictive weather data via the Open-Meteo API using exact geocoded town/city coordinates.
- Machine Learning Prediction Engine: Utilizes a Random Forest Classifier trained on 3,000+ synthetic climatic data points to calculate a Disease Risk Score in milliseconds.
- In-Memory Knowledge Graph: Features a custom Python dictionary structure that instantly fuzzy-matches crop types to verified prevention and treatment protocols, eliminating database latency.
- Calamity Alert System: Actively scans forecast arrays to trigger visual UI warnings for impending extreme weather events like heavy rainfall.
- Interactive Dashboard: Dynamic mixed-chart trend visualizations (Lines for Temp/Humidity, Bars for Rain) built natively in the browser.
The project utilizes a decoupled hybrid cloud infrastructure for maximum scalability and low-latency global delivery.
| Component | Technology | Description |
|---|---|---|
| Backend API | Python, FastAPI | Asynchronous data fetching, ML inference, and API orchestration. Hosted on Render. |
| Machine Learning | Scikit-Learn, Joblib | Random Forest model (agrisense_model.pkl) mapping climatic thresholds. |
| Frontend UI | HTML5, CSS3, Vanilla JS, Chart.js | Lightweight, interactive client. Hosted on Vercel Edge Network. |
AgriSense-AI/
├── .gitignore
├── agrisense_model.pkl
├── index.html
├── main.py
├── requirements.txt
└── train.py