Climate Risk Scoring API

Know the risk before you sign the loan.

Drishti fuses satellite data, weather stations, climate models, and on-the-ground sensors into a single, property-level climate risk score — with honest uncertainty bounds. Built for insurers, lenders, and platforms that can't afford to guess.

Request Early Access

The problem is real

Climate risk data today is fragmented, opaque, and often overconfident. Decisions worth millions get made on incomplete pictures.

$120B

Insured losses in 2024

The fourth year in a row above $100 billion. The old actuarial models aren't keeping up with the pace of change.

40%

Properties mispriced for flood

FEMA maps are outdated and binary — you're in or out of the zone. Reality is a gradient, and lenders need to see it.

Disconnected data sources

Satellite retrievals, ground stations, reanalysis products, sensor networks, climate projections — nobody's fusing them properly.

How Drishti works

We use Bayesian Maximum Entropy data fusion — the same geostatistical framework used in peer-reviewed air quality research — adapted for multi-hazard climate risk.

Ingest

We pull from NOAA stations, FEMA floodplains, satellite-derived moisture, elevation models, and CMIP6 climate projections. Hard data and soft data, treated differently.

Fuse

Our BME engine merges these sources at the property level, weighting each by its actual information content — not just averaging them together.

Score

You get a composite risk score plus individual hazard scores (flood, heat, wildfire, wind) — each with a confidence interval so you know what you're working with.

Query

One API call. Send coordinates or an address, get back a JSON response with scores, uncertainties, and contributing hazard breakdowns.

Simple API, rich output

A single endpoint. Coordinates in, risk scores out — with the uncertainty bounds your actuaries actually want.

      // GET /v1/risk?lat=29.76&lng=-95.36

      {

        "address": "Houston, TX 77001",

        "composite_score": 78,

        "confidence_interval": [72, 84],

        "hazards": {

          "flood": { "score": 91, "ci": [85, 96] },

          "heat": { "score": 67, "ci": [58, 74] },

          "wildfire": { "score": 12, "ci": [8, 19] },

          "wind": { "score": 65, "ci": [55, 73] }

        },

        "data_sources": 7,

        "assessment_date": "2026-04-07"

      }

Who this is for

If you price risk, underwrite assets, or make decisions tied to physical locations, Drishti gives you better numbers.

Insurance Underwriting

Move beyond FEMA zones and catastrophe models. Get property-level, multi-hazard scores with honest uncertainty to price policies that reflect actual exposure.

Mortgage & Lending

Screen loan portfolios for hidden climate concentration risk. Regulators are asking for it. Your models should already include it.

Real Estate Platforms

Add a climate risk layer to property listings. Buyers increasingly want to know what they're walking into — and disclosure rules are tightening.

ESG & Disclosure

Physical risk scoring for TCFD and SEC climate disclosure requirements. Asset-level, auditable, with methodology you can actually explain to stakeholders.

Built by someone who's done this before

I'm Praful Dodda, a PhD candidate in Environmental Sciences and Engineering at UNC Chapel Hill. My research under Prof. Marc Serre focuses on Bayesian Maximum Entropy data fusion for air quality — the same statistical framework that powers Drishti's climate risk engine.

My dissertation spans three published studies fusing satellite data, chemical transport models, and ground monitoring networks across national and global scales. Before Drishti, I placed in the top 2% (12th of 664 teams) in NASA's 2023 Air Quality Forecasting Challenge.

Drishti isn't a pivot into a hot market. It's the direct application of six years of spatiotemporal data fusion research to a problem that desperately needs it.

UNC Chapel Hill PhD BME / Geostatistics NASA Challenge Top 2% AWS Certified Python + HPC Multi-Source Data Fusion