Insure all water risk.

Water is the largest insured-loss peril of the next decade. Quantile Labs is building the AI that prices it, from sensors in the catchment to capital on the treaty.

OUR GOAL

Making water risk insurable.

Climate change is rewriting the loss distribution faster than any model in the market can keep up. Quantile Labs is the system that catches up.

Hydrologists, catastrophe modellers, and machine-learning specialists, shipping the science as the AI underwriters and reinsurers call.

SIGNAL

Sediment plume, river stage, reservoir colour. The catchment, observed from orbit and instrumented from the ground.

QUANTILE

Our name is the statistical object reinsurance is priced on: the quantile of the loss distribution. The 1-in-200 tail.

By modelling the full distribution rather than the average, the lab prices the events that move portfolios.

Our Tech

Three calls. One water-risk AI.

Underwriting, capital, and event response on one calibrated surface.

Score any flood-exposed address in seconds.

Sub-metre water-risk score for any postcode, calibrated against owned sensors and satellite extents. One surface, four perils, returned in a single response.

  • Peril scores
  • Depth-damage
  • First-floor priors
  • Contents & BI

Price the treaty against the full distribution.

OEP, AEP, TVaR, and capital at risk for any layer attaching at any return period. The full quantile surface, returned by construction.

  • OEP / AEP
  • TVaR · 1-in-200
  • Treaty optimisation
  • IFRS 17 ready

Reconstruct event footprints within hours.

Owned-sensor telemetry and satellite imagery fused into a calibrated footprint. Portfolio impact and treaty triggers on the same call.

  • Footprint reconstruction
  • Sensor & satellite fusion
  • Portfolio impact
  • Trigger calls
SURFACE

Depth, return period, exceedance — the full quantile of loss returned directly by construction.

Four properties of the water-risk AI.

  1. I

    Physics before precedent.

    Models begin from rainfall-runoff physics and shallow-water dynamics. Fitting comes after.

  2. II

    Ground truth, owned.

    The lab operates its own gauges, rain, water-level, and soil-moisture sensors. Calibration runs on telemetry the lab owns.

  3. III

    Decision-grade uncertainty.

    Every output ships with provenance, assumptions, and the full quantile surface, in the units reserving teams already use.

  4. IV

    Science that ships.

    Every program graduates from notebook to a callable surface. The lab's output is software.

OUR HERITAGE

Hydrology has decades of physics. Quantile builds on it.

Our Tech
BACKED BY

Backed by the institutions building tomorrow's research.

  • LSE Generate
  • Imperial College London
  • London School of Economics
CONTACT