See climate losses months before they happen.

ClimateExpert is a geospatial AI platform that forecasts climate events months ahead, asset by asset, on a 25 × 25 km grid: droughts 1–3 months out with 90% accuracy, floods 1–6 months out with 88%. Every forecast is published in advance, then shown next to the fact.

In production at a major European bank ERA5 · Copernicus · 336 satellites Forecasts published before the season, checked after Deploys in your closed contour
$417B
estimated global economic losses from climate events in 2024
Source · Gallagher Re
$154B
global insured climate losses paid out in 2024
Source · Gallagher Re
+8–9%/yr
growth in hazardous weather events since 2007. Historical loss tables no longer price the future
Source · WMO

Climate risk has outgrown the tools.

The past no longer prices the future

Hazardous weather events grow 8–9% a year. Loss tables built on history systematically misprice today's risk.

Regulators require hard numbers

ECB & EBA climate stress tests, IFRS 9 provisioning, Solvency II and ORSA all demand quantified physical climate risk.

Decisions live on a 12-month horizon

Scenario platforms describe the year 2100 for disclosure. Underwriting, renewals and credit decisions happen this season. ClimateExpert operates exactly there: 1–12 months ahead.

Three minutes inside the platform.

Portfolio on the map, per-asset probabilities, alerts and one-click reports, recorded on the live system.

Interface preview · live demo available on request

What the underwriter sees.

Upload your assets: sites, fields, plants, pipelines. The platform answers the questions a risk team actually asks.

  • Which assets fall into the event zone, and where risk concentrations form across the book.
  • Which assets need extra inspection, who should receive a warning, and where a large aggregate loss is possible.
  • When the picture changes: forecasts recompute as new station, radar and satellite data arrives, and alerts fire past your threshold.
  • What to hand to the board: CSV, XLSX and PDF exports with map and risk tables, ready for management, reinsurers and regulators.

Eight things your team can do in the first quarter.

Price climate into every policy

ML probabilities per asset for each hazard, with traffic-light scoring and differentiated tariffs by risk level.

See the whole book

Portfolio map with geographic risk concentrations and probable maximum loss for catastrophic scenarios.

Enter renewals with your own seasonal view

Seasonal outlooks moved cat reinsurance pricing by 15–20% at the 2026 mid-year renewals (TSR). Hold an asset-level signal and justify your reinsurance structure to the board and the regulator.

Verify claims from the map

The platform confirms the natural event and shows the insured asset inside the impact zone, automating first-pass claims verification.

Act on change

Forecasts recompute as new data arrives. Alerts fire when an asset's risk shifts past your threshold, in-app and by email.

Report in one click

CSV, XLSX and PDF exports for a region or a portfolio. The PDF includes the map and risk tables, a ready document for the board or the regulator.

Move climate into credit decisions

The same engine screens investment projects for climate risk at one of Eastern Europe's largest banks.

Keep every byte inside

On-premise deployment with integration into internal systems, role-based access and team training. Portfolio data never leaves your infrastructure.

Accuracy is the contract.

Forecasts are published in advance and later shown next to the fact. The free month includes a backtest on your own loss history.

Soil drought, region A

Agricultural region · six weeks of lead time
Forecast · 16 May 2026SMI −0.61
Fact · 30 June 2026SMI −0.47

Drought materialised where predicted, six weeks after the forecast was published.

Soil moisture, region B

500 km away · same forecast run
Forecast · 16 May 2026SMI +1.40
Fact · 30 June 2026SMI +1.04

No drought, as predicted. Opposite outcome to region A, both correct.

Conventional weather forecastAll providers
~14 days
Droughts · ready90% accuracy · field-tested with a national agri-insurance association
up to 3 months
Floods · ready88% accuracy
up to 6 months
Storms & wildfiresIn development · target 80%+
up to 12 months
today2 mo4 mo6 mo8 mo10 mo12 mo

Also on the map: permafrost degradation and earthquakes as historical-event layers with probabilistic hazard assessment. Models run on ERA5 / Copernicus / GFS, radar and station data, with EM-DAT, CRED and NOAA event records, using ensemble and deep-learning methods.

30 daysahead of the flood

Our 2024 seasonal flood forecast was published 30 days before a major river flood and confirmed by satellite imagery: street-level flood extent and water depth, exactly where predicted.

Foresight has a market price.

$1 → $10

Each dollar invested in early warning saves up to ten in avoided losses.

15–20%

How much seasonal outlooks moved catastrophe reinsurance pricing at the 2026 mid-year renewals.

$16M

Avoidable losses at a single dam in a single year under forecast-based drought operation.

$323M/yr

The value of seasonal ENSO prediction to US agriculture alone.

Your number depends on your book. The result of applying the technology: smoother financial results through catastrophe scenarios, optimised reinsurance spend, stronger capital adequacy, and the ability to work in regions once considered too risky.

Public tiers. The first month is free.

The market sells climate analytics through opaque enterprise quotes. Our tier structure is open, and the trial costs nothing.

Trial

Free · 1 month
  • Full web platform in the browser
  • Forecast archive for the 2024–2025 seasons
  • Historical catastrophe layers: earthquakes, permafrost
  • Dashboards and analytics for any area on the map
  • Hazard classification guide
  • Accuracy backtest on your own loss history
  • No company data enters the system
Start a free month

Standard

Pricing on request
  • Everything in Trial
  • Live forecasts across the coverage area
  • Monthly forecast updates
Request a quote

Expert

Pricing on request
  • Everything in Standard
  • Portfolio upload to the map
  • Automated alerts on risk-profile change
  • XLSX export by grid cell and region
  • On-premise deployment in your closed contour
  • Integration with internal systems
  • Role-based access for staff
  • Team training and implementation support
Request a quote

Quotes depend on portfolio size and deployment. Custom hazard layers, for crops or building types, are scoped individually.

Put your portfolio on the map.

The first month is free and includes an accuracy backtest on your own loss history. No company data enters the system.