Climate risk modeling and insurance underwriting with machine learning
Kettle applies machine learning to quantify climate-driven catastrophic risk and underwrite policies against it. The stack is cloud-native (AWS Lambda, ECS, RDS, PostgreSQL) with strong infrastructure-as-code discipline (Terraform, CloudFormation, CDK), and current project focus—CI/CD, deployment reliability, observability—signals an engineering organization scaling from early product-market fit into operational rigor. All six active hires are senior-level engineering and product roles, suggesting they're building depth rather than breadth.
Kettle underwrites insurance and reinsurance policies for climate-driven catastrophic risks like wildfires and hurricanes, using machine learning to model and price exposure. The company operates from Lemoore, California with a small, senior-focused team of 11–50 employees. Revenue model flows from direct premium underwriting and reinsurance transactions. Current work spans wildfire risk modeling, expanding peril coverage (beyond wildfire into hurricane and other perils), and a new policy administration platform—each requiring both actuarial domain work and backend infrastructure maturity.
AWS services (Lambda, ECS, RDS, SNS), PostgreSQL, Python backend, React/TypeScript frontend, Terraform and CloudFormation for infrastructure-as-code, Datadog and CloudWatch for observability, Docker for containerization, and SAS for actuarial modeling.
Infrastructure-as-code rollout, CI/CD and deployment reliability, observability and monitoring, a new policy administration platform, and expanding wildfire and property risk modeling capabilities.
Other companies in the same industry, closest in size