This project provides a comprehensive, web-based platform engineered to enhance disaster preparedness, response, and mitigation against five critical natural hazards: floods, forest fires, hurricanes, earthquakes, and droughts.

​Built on a modular Flask backend with MongoDB for secure data persistence and user management, the system integrates tailored Machine Learning models to deliver reliable risk predictions. The predictive core features an ensemble model (Decision Tree, Random Forest, KNN, XGBoost) for flood forecasting, while optimized Random Forest classification and regression models handle fire, hurricane, earthquake, and drought risk calculations using environmental and spatial inputs (e.g., atmospheric pressure, seismic depth, temperature, elevation, monsoon trends).

​The platform prioritizes accessibility during emergencies with a responsive dashboard, interactive mapping (Leaflet/Mapbox), real-time alert systems with multi-level risk categorization, and support for 5 distinct languages. Designed to scale for high-traffic emergency scenarios, it bridges complex ML analytics with intuitive visualizations, giving citizens and emergency authorities actionable intelligence when time matters most.