HarperYC W25

A bar owner gets covered before lunch

A daycare in Ohio needs a certificate of insurance by Friday. A car dealership in Texas is up for renewal. A welding shop wants workers' comp without a week of phone tag. In the old world, each of these is a small ordeal. At Harper, each is a job that mostly runs itself - an AI fills the forms, finds the right carrier among 160-plus, chases the underwriter, and comes back with a quote while the owner is still on their first coffee.

That is Harper today: a licensed, near-autonomous commercial insurance brokerage out of San Francisco, insuring more than 5,000 businesses across 35 states. It looks less like an agency and more like a software company that happens to hold an insurance license. The boring part of running a business - making sure a disaster doesn't end it - finally moves at the speed of everything else.

Insurance always got in the way

Commercial insurance is a $100B-a-year commission market that still, somehow, runs on PDFs, email threads, and human memory. Applications are long. The information is asymmetric. A single broker can realistically manage 20 to 30 deals a month, because each one is a slog of submissions and follow-ups. The result is a customer experience nobody would design on purpose and a whole layer of businesses - the daycares, the welders, the local restaurants - who get treated as too small to bother with.

Harper's founders did not learn this from a market report. CEO Dakotah Rice grew up in rural Alabama, watching his father run a nightclub and his uncle's trucking company struggle. The recurring villain in those stories was paperwork with a deductible attached.

Two ex-Goldman builders, one second swing

Harper is not Dakotah Rice and Tushar Nair's first rodeo. The two previously built Poolit, a fintech that reached around $100M in assets, and both did time at Goldman Sachs - Rice on the investing side, with later stops at Carlyle and Coatue; Nair leading machine-learning engineering. Their first idea here was to sell AI software to existing brokers. It didn't work. So they did the harder, more interesting thing: they got licensed and became the brokerage themselves.

The bet is simple to state and hard to pull off. If AI can absorb the repetitive machinery of a brokerage - routing, forms, document chasing - then a small team can serve thousands of clients instead of dozens, with better service and better margins. The AI learns from the brokers' own expertise, accumulating the workflows and judgment that, over time, point toward genuine autonomy.

From a failed pivot to a record round

Harper Hub does the unglamorous work

At the center sits Harper Hub, the AI engine that handles the parts of brokering that no one enjoys. It completes the application forms. It identifies the optimal carrier and underwriter for a given risk. It routes submissions, manages follow-ups, collects documents, and keeps client communication moving. Harper places workers' compensation, general liability, and professional liability for small and mid-sized businesses - the unsexy coverages that keep a company's doors open.

The coverage is real and licensed; the difference is throughput. Where a traditional broker juggles a few dozen deals a month, Harper's system handles more than a thousand. That is the whole argument in one number.

Caption: somewhere, a stack of unfilled ACORD forms is quietly weeping.

The money showed up

In February 2026, Harper announced $47M in combined seed and Series A funding, led by Emergence Capital, with Y Combinator, Peak XV Partners, Lobster Capital, Antler, 10X Founders, Fellows Fund, and Outset Capital along for the round. It was reported as the largest publicly disclosed Series A ever raised by a Black founder in the United States - a headline Rice would probably trade for a few thousand more renewals, but a milestone all the same.

The traction underneath it is the more convincing part. In roughly 13 months, Harper served over 5,000 businesses - manufacturers, healthcare providers, hospitality groups, restaurants - and wrote more than $6M in annualized premiums since switching customer operations on in October. The carrier roster reads like the actual market: AIG, GEICO, Hiscox, James River, USLI, and others.

Largest brokerage, best NPS - in that order

Harper's stated goal is cheerfully immodest: become the largest insurance brokerage in the world, with the best net promoter score. The near-term focus is mid-sized businesses with complex needs; the long-term vision is to serve every business in the country and, eventually, to become the place owners go for anything touching risk, compliance, and back office.

What makes that more than a slide is the order of the ambition. Plenty of companies want to be the biggest. Wanting to be the biggest and the most liked, in an industry famous for being neither, is a specific kind of dare.

Back to the bar owner

Return to that daycare in Ohio, that dealership in Texas, that welding shop. In the world Harper is betting on, none of them spends a week on hold to protect what they've built. The certificate arrives before Friday. The renewal quotes itself. The owner gets back to the actual work.

That's the quiet promise underneath the AI and the $47M: insurance that stops getting in the way. Harper hasn't finished proving it can do that at the scale of "every business in the country." But it has changed the opening scene already - and for a few thousand businesses, the paperwork purgatory is now just an afternoon.

Profile compiled from public sources · Figures are company-stated and approximate