The Insurance
Startups
Actually Want

Right now, somewhere, a startup founder is sitting in a waiting room they didn't ask to enter.

Their VC said they need a certificate of insurance before the term sheet closes. Their broker said it'll take two weeks. Their legal counsel said "make it happen." And their co-founder is about to post their launch tweet.

This is the moment Corgi was built for. The San Francisco startup - founded in 2024, unicorn by 2026 - has quietly become the default insurance solution for the startup ecosystem. Not by writing clever copy, but by making the two-week process a five-minute one.

Corgi is not a broker. It is not a middleman. It owns its own carrier, underwrites its own policies, and prices risk through an AI system called Hammurabi - named, fittingly, after one of history's first lawgivers. Forty thousand startups across 49 U.S. states now use it. Their average annual premium is around $1,000. Their churn rate is less than one percent.

When Emily Yuan says "centuries ago," she means it literally. The Lloyd's Coffee House model - syndicates of underwriters assessing risk over candlelight - dates to 1688. The paperwork has been digitized since then, but not much else. Corgi changed that by building the whole stack themselves, from pricing algorithms to claims processing, and asking 40,000 startups if they'd like to try it. They said yes, and they stayed.

Every startup needs insurance. Almost none of them want to buy it.

Business insurance is the regulatory tax that comes with building a real company. Investors require it. Enterprise customers demand it for contracts. Regulators mandate it to operate. The problem is not whether you need it - it's how the industry has always made you get it.

Traditional commercial insurance runs on a distribution chain with at least three middlemen: the retail broker who collects your information, the wholesale broker who packages it, and the carrier who decides whether to write the policy. Information flows up slowly. Quotes come back in days or weeks. Policies are bespoke documents that bear little relationship to how a modern AI startup actually operates.

Nowhere is this gap more obvious than in AI liability coverage. As of 2025, most legacy carriers exclude AI-caused errors from their tech E&O policies entirely - or write exclusions so broad they're functionally useless. For a startup whose entire product is an AI system, that coverage gap is an existential risk dressed up as a policy document.

Corgi was built to fill that gap - and then to obliterate the process that created it.

Two founders who built games for 200 million people decided to fix insurance instead.

They met in college. The pivot from consumer social apps to insurance sounds improbable until you understand their actual insight: they weren't trying to become insurance professionals. They were trying to build software that replaced what insurance professionals did manually.

The Y Combinator Summer 2024 batch accepted Corgi in the same cohort that produced dozens of AI startups. Within 7 months of graduating, the company had raised $108 million. Within 11 months, it was worth $1.3 billion. The speed is unusual, even by YC standards. YC itself called Corgi its latest unicorn.

The proof is in the retention.

A sub-1% churn rate in insurance is not a marketing number. It means customers who bought Corgi's product stayed. In a market where buyers grudgingly re-shop at renewal every year, that kind of retention signals something structural: the product is better, and the switching cost is no longer just inertia.

Customers include Deel and Artisan - companies that have options. Average revenue per customer sits around $1,000 annually, which makes the unit economics unusually clean: a high volume of small policies, automated at scale, with minimal acquisition cost inside the startup ecosystem.

Every important company eventually needs insurance. Corgi is building who they call.

The startup insurance market is large but finite. Corgi knows this. The Series B announcement came with an explicit move into trucking - a $50+ billion commercial insurance vertical that looks, in many ways, like startup insurance did before Corgi showed up: high friction, slow quotes, commodity coverage that doesn't match actual risk profiles.

Payroll insurance and small business coverage are on the roadmap. New offices are planned in Salt Lake City, Atlanta, Chicago, Dallas, New York, and London. The company is 140 people as of mid-2026 and adding.

What Corgi is building - beneath the startup-friendly branding and the orange corgi mascot - is a modern insurance infrastructure company. The premise is that if you own the carrier, own the underwriting algorithm, own the claims process, and own the distribution, you can offer better products at better prices and still capture better economics than the traditional chain.

The theory is not new. What's new is that it seems to be working. The 40,000 startups, the $40M ARR, the sub-1% churn, the 18-month journey from YC batch to unicorn - these are not projections. They're receipts.

The details that make Corgi, Corgi.