Reask GmbH

Metryc, DeepCyc

Reask applies machine learning and climate science to tropical-cyclone risk through Metryc (event response and footprints) and DeepCyc (forward-looking, climate-conditioned stochastic event sets), modelling wind and rainfall hazard globally and back through the historical record. Its outputs act as data and calculation inputs for pricing and triggering parametric (re)insurance contracts, with climate-conditioned views of how cyclone risk shifts under warming. Distinctive as a cyclone specialist embedded in the parametric market, focusing narrowly on tropical-cyclone hazard rather than multi-peril coverage.

Vendor methodology

Transparency Score

Beta

Public transparency, not model quality iThe Transparency Score (0–3) estimates public methodological transparency: the degree to which the vendor's analytical approach to modeling climate and nature risk can be assessed from publicly available sources. It is explicitly not a measure of vendor quality or accuracy.

2

Methodology-relevant: vendor websites or external resources explain methodological details.

Methodology-relevant: public product and API documentation name the model components reconstructably — DeepCyc uses ERA5 reanalysis + NCAR-CESM ensembles for stochastic catalogues with teleconnection sensitivity (ENSO/AMO/IOD) and 1 km boundary-layer physics; Metryc validated against 1,200+ anemometer observations on public agency data (NHC/JTWC/BOM). Full white paper brochure-gated; parameters proprietary.

For further information, see the section 'Transparency Score methodology' in the Readme.

Contact us to point out or add relevant source or references.

Links

i

Level 1 signals