Corvus Insurance

Commercial cyber insurance emerged as one of the fastest-growing specialty lines in the early 2010s, yet the market lacked the underwriting sophistication the risk category demanded. Traditional insurers relied on manual processes and generic risk frameworks ill-suited to assessing dynamic, organization-specific cyber exposures. The result was widespread mispricing — carriers either left profitable risk on the table or accepted policies they couldn't adequately evaluate, driving poor loss ratios across the segment. Without proprietary data or quantitative models, incumbents had little ability to differentiate high-quality risks from high-hazard ones, limiting both market participation and profitability in a rapidly expanding line.

Corvus built a proprietary technology platform that applied predictive machine learning to cyber insurance underwriting from the ground up. Rather than layering AI onto legacy systems, the company embedded data science and automation into the core underwriting workflow — ingesting unique datasets to generate more granular risk signals than traditional questionnaire-based processes could provide. This enabled automated risk scoring and pricing decisions calibrated to each insured's actual cyber posture. The approach combined deep insurance domain expertise from founder Phil Edmundson with a modern data infrastructure guided in part by strategic input from investors, including technical guidance on the technology stack and the recruitment of a founding CTO and CISO to lead the build-out.

Corvus's ML-driven underwriting delivered a sustained, measurable edge over the broader market:

• 15–20 percentage point outperformance versus industry loss ratios, consistently maintained across underwriting cycles
• Gross premium written scaled from a few hundred thousand dollars to over $275M annually — growth achieved without sacrificing underwriting discipline
• Acquired by Travelers, one of the top three U.S. insurers, in the largest insurance exit of the year and the 5th largest in the last decade

Beyond the financials, Corvus became Travelers' platform for entry into the cyber insurance market, validating the strategic value of the underwriting model itself.

• Proprietary data compounds over time: in specialty lines, the insurer that builds the best risk dataset earliest creates a durable moat that generalist carriers cannot close quickly.
• Domain expertise is not optional: combining insurance underwriting knowledge with data science — rather than treating them as separate functions — was central to Corvus's accuracy advantage.
• Embed AI in the workflow, don't bolt it on: building ML into core underwriting decisions from day one produced better loss ratios than retrofitting automation onto existing processes would have.
• Strategic talent introductions can be as valuable as capital: early recruitment of a technically strong CTO and CISO shaped the platform's architecture at a formative stage.