EN
Back to the archive

The archive · Money & Fintech · Strategic decision · 2024-2026

Harper rebuilt the insurance brokerage around AI, won 5,000 clients, and raised $46.8M

YC W25 agency that owns the whole insurance workflow with AI, not just tools; it handles 1,000+ customers a month and raised $46.8M from Emergence.

Harper

The betThat AI could do the judgment-heavy work of a commercial insurance broker well enough for an AI-run agency to out-execute fragmented incumbents and scale.Scaling

What the business is

AI-native commercial insurance brokerage that matches small and mid-sized businesses with 160+ carriers for workers' comp, general, and professional liability — it owns the agency, not just software.

Starting capital$46.8M combined seed and Series A, led by Emergence Capital, with YC, Lobster Capital, and Peak XV (Feb 2026).

How it started

Dakotah Rice, a Goldman Sachs and Carlyle alum whose family owned an insurance brokerage, had just closed his fintech Poolit in 2023 and admitted he should have shut it down a year earlier. With Tushar Nair, Poolit's former CTO, he first considered building AI tools for existing brokerages — then decided the tools were the easy half and the better bet was to be the brokerage itself. Harper launched in 2024 and joined Y Combinator's Winter 2025 batch.

What happened

Harper took its first customers in October 2024 and by its YC launch was writing over $6M in annualized premiums across 35 states. AI handles submission routing, underwriter follow-ups, document collection, and pipeline management, while proprietary models match each business against Harper's network of 160+ carriers, wholesalers, and MGAs. The result, per the company: what takes a traditional broker five to seven days gets done in one to two, and a sales engine that handles 1,000+ customers a month versus the usual 20-30. By February 2026 Harper said it had more than 5,000 customers.

How it ended up

Still active and scaling. In February 2026 Emergence Capital led a combined $46.8M seed and Series A with YC, Lobster Capital, and Peak XV, and Harper was hiring to grow its engineering team and its brand in middle America — daycares, manufacturers, dealerships, bars, and restaurants.

Background

Harper is a commercial insurance brokerage built around AI rather than an old agency with AI added. Dakotah Rice — whose family owned a brokerage and whose uncle's trucking company was crushed by insurance costs — and Tushar Nair, a former Goldman engineer, first thought about selling AI tools to existing agencies. They decided instead to become the agency, launching Harper in 2024 and going through Y Combinator's Winter 2025 batch.

The bet was that AI could take over the judgment-heavy work of brokering: reading applications, routing submissions, following up with underwriters, and matching each business against carriers. Harper says its AI works across 160+ carriers, wholesalers, and MGAs, cutting a five-to-seven-day process to one or two days and handling more than 1,000 customers a month where a human team handles 20 to 30.

By February 2026 Harper said it had 5,000+ customers, and Emergence Capital led a combined $46.8M seed and Series A with YC, Lobster Capital, and Peak XV. The company's ambition is to become the largest brokerage in the world and eventually the back office for small-business owners — insurance first, then risk, compliance, and beyond.

What has to be true

  • Owning the service captures the data every policy produces, and that data feeds the matching model — the work compounds instead of leaking to clients.
  • Incumbents run fragmented, manual workflows; an AI-native agency can serve niches the old model cannot scale to profitably.
  • Rice's family brokerage background gave him the customer and the credibility to enter a relationship-driven industry.
  • The Poolit failure taught him to cut losses earlier and gave the second company a sharper thesis.

What can be applied

When AI makes a service scalable, the strongest position is to own the service, not sell the tool: Harper's data compounds with every policy, and the customer relationship stays in the house.

Aftermath

As of February 2026, Harper had 5,000+ customers, was handling more than 1,000 customers a month, and was using the $46.8M round to grow its engineering team and its brand in middle America — daycares, manufacturers, car dealerships, local bars, and restaurants. The founders' stated endgame was to become the voice for entrepreneurs: insurance first, then risk, compliance, and the rest of the back office.

Sources

spotted an error? The archive wants to know.

Your turn

You just read one. Describe what you are building, and see who is betting on the same thing.

Free account · 3 free questions · no card

Related cases