SurgeShield uses machine learning to analyze environmental data — rainfall, river levels, elevation, soil type — and predicts flood risk in real-time. Protecting communities, saving lives.
People affected by floods globally between 1998 and 2017 — more than any other natural disaster.
Annual global flood damage to homes, infrastructure, and livelihoods.
SurgeShield's prediction accuracy, validated on 2,000 held-out flood observations.
Data: WHO · World Bank · SurgeShield ML Pipeline
Three simple steps from data to protection
Enter rainfall, temperature, river discharge, elevation, soil type and other parameters for your location.
Our XGBoost model processes your data against 10,000 training observations to classify flood risk.
Receive flood probability, risk level, and the top contributing factors driving the prediction.
Every capability built for clarity, speed, and trust.
12-parameter input form with intelligent defaults and real-time ML prediction served by a production API.
Visualize flood-prone zones with color-coded markers across 10,000 observations on a dark-themed Leaflet map.
Compare 5 ML models, view confusion matrices, ROC curves, and feature importance — all in one place.
Citizen monitoring and administrator oversight, secured with Clerk authentication.
Built on publicly available environmental datasets.
Every prediction explains which factors drove the result.
Model trained on universal physics — works across geographies.
HTTPS, VPS, Nginx — not just a research prototype.
Start predicting flood risk for free with SurgeShield.