Kettle builds a reinsurance underwriting platform that uses machine learning, weather data, satellite imagery and other datasets to model catastrophic risk, starting with California wildfires. The company sells reinsurance technology to customer-facing insurers and also resells policies through one channel aimed at very high-value properties (roughly $3M–$10M). Kettle says its models run over 500 billion simulations and have been accurate about 89% of the time so far. The startup has engaged roughly 26 insurance carriers for modeling and expects three to four commercial deals to close by the end of the year. Longer term, Kettle plans to expand its tools to additional disaster types and other states. Kettle is a San Francisco-based reinsurance startup that uses advanced deep learning and proprietary algorithms to protect people from climate-change-related risks. Its first product provides wildfire reinsurance to protect Californians’ businesses, homes, and livelihoods. The company leverages swarm-intelligence machine learning and terabytes of public and private data sources, including NOAA weather data and NASA’s MODIS and LIDAR satellites, to model fire risk. Kettle’s neural networks run upward of 140 million parameters and calculate probabilities of fire damage at half-square-mile resolution across California. In 2020 the company’s model flagged all fourteen of the largest wildfires among the top 20% of areas most likely to burn, a performance cited in the article. Kettle raised $4.71M in seed financing led by True Ventures with participation from Acrew Capital, Homebrew Ventures, Anthemis, and Inspired Capital.

Total raised

$30M

Funding rounds

2

Latest round

Series A

Latest activity

Nov 2021