Data in. Intelligence out. Nothing hidden in between.

One continuous record, not a pile of disconnected extracts.

Observed climate, satellite and remote sensing, terrain, hydrological records and climate projections are held as a single, consistent series — the same record for 1951 as for last monsoon — so trends are real trends, not artefacts of a change in source.

• Observed rainfall and river levels
• Recorded historical flood events
• Satellite and remote sensing observations
• Terrain, drainage and land use
• Climate projections for future scenarios

The right model for each question — not one model for everything.

Geospatial processing, statistical analysis, hydrological reasoning and machine learning are each applied where they are strongest. Machine learning earns its place where it measurably improves a result; it is never used to paper over gaps in the data.

• Geospatial processing at district and taluka scale
• Statistical and hydrological modelling
• Machine learning where it earns its place
• Reproducible, versioned model runs

Written down, so a result can be argued with rather than taken on trust.

Risk is built from three parts — the hazard (how often and how badly an area floods), the exposure (what is in the water's path) and the vulnerability (how much damage it would do). How they are combined is documented, with every assumption stated.

• Hazard × exposure × vulnerability, documented
• Assumptions stated, not implied
• Uncertainty reported with every figure
• Versioned method notes

Checked against the floods that actually happened.

Model output is back-tested against recorded flood events, and those checks are reported alongside the result — not kept in a separate file nobody reads. Where the model is less certain, we say so.

• Back-testing against observed floods
• Independent review of method
• Published performance notes

Intelligence that fits the workflow you already run.

Use FloodGuard through Explorer, Atlas and Live, or bring the same intelligence into your own underwriting, lending, GIS or planning systems — instead of moving your team into ours.

• Dashboards
• REST endpoints
• Bulk data
• Map tiles
• Reports