Kalepa for Specialty Lines Underwriting

In-Depth Review

Kalepa was founded in 2018 and went through Y Combinator, positioning itself as an AI copilot specifically for specialty and excess lines underwriting. While most underwriting AI tools target the broad commercial market, Kalepa focuses on lines where underwriting complexity is higher and the consequences of appetite drift are more severe: workers compensation, commercial auto fleet, and professional liability.

The Copilot Approach

Kalepa positions itself as a copilot rather than an automation engine. The distinction matters. The platform does not make underwriting decisions. It reads submission documents, extracts risk characteristics, checks them against the carrier’s appetite rules, and presents its findings to the underwriter with a recommendation. The underwriter still makes the call.

This is a deliberate design choice for specialty lines. In workers compensation, for example, the difference between a profitable account and a loss-ratio disaster can hinge on details that AI models are not yet reliable enough to evaluate independently: the quality of the insured’s safety program, the experience of their return-to-work coordinator, or whether their loss history reflects systemic problems or one-time events.

Appetite Guardrails

The appetite guardrail feature is where Kalepa provides its most concrete operational value. In specialty lines, underwriting guidelines exist as documents that underwriters are expected to follow. In practice, production pressure and relationship dynamics lead to appetite drift: underwriters bind risks they know are marginal because a broker relationship matters or because the submission arrived on a slow month.

Kalepa makes appetite rules enforceable. If a submission falls outside defined parameters (class codes, territory, loss ratios, premium size), the system flags it and requires a management override before the underwriter can proceed. This is not optional compliance guidance; it is a workflow control.

Loss Run Analysis

Loss run analysis automation is Kalepa’s second core feature. In experience-rated lines like workers compensation, the loss run is the single most important document in the submission. Reading loss runs manually is tedious: they arrive in inconsistent formats from different carriers, often as multi-page PDFs with varying layouts.

Kalepa reads loss runs, extracts claims data, calculates experience modification factors, and identifies trends in claim frequency and severity. For underwriters handling dozens of experience-rated accounts per week, this automation removes hours of manual data extraction.

Limitations to Consider

Kalepa’s specialty focus means carriers writing general commercial (BOP, CGL, commercial property) will not find applicable models. The platform is also smaller than competitors like Federato or Sixfold, with a more limited deployment base. This means fewer carrier references and less production-validated model performance data.

The copilot model requires experienced underwriters who can evaluate the AI’s recommendations critically. Organizations looking for straight-through processing or fully automated underwriting will find that Kalepa assists the underwriter rather than replacing underwriter judgment.

Who Should Evaluate Kalepa

Specialty carriers and MGAs with 10 or more underwriters writing workers compensation, commercial auto fleet, or professional liability. The appetite guardrail value is highest in organizations where underwriting guidelines have been inconsistently enforced, and the loss run automation value is highest in experience-rated lines with heavy submission volumes.

Request a proof-of-concept using your own loss runs and appetite guidelines. Evaluate the document extraction accuracy on your specific loss run formats, and confirm that the appetite rule configuration matches the nuance of how your guidelines actually work in practice.