A Typical Brokerage Does 30 Deals a Month. Harper Does 1,000.

01 /The Concept

Commercial insurance has thousands of brokerages and almost none of them can process more than a few dozen deals a month. That isn't a demand problem – it's a workflow problem. Submission routing, underwriter follow-up, document collection, pipeline management: every step is done manually by a person who can only do one thing at a time.

Harper is an AI-native licensed commercial agency that does all of it automatically. Founded in 2024 by Dakotah Rice and Tushar Nair – who had shut down a prior startup together – Harper went through Y Combinator's W25 batch and raised $46.8 million in February 2026 in a combined round led by Emergence Capital. The company matches small and mid-sized businesses with more than 160 insurance carriers across workers' compensation, general liability, and professional liability lines, and places coverage in 24 to 48 hours. A traditional brokerage team takes five to seven days.

The throughput numbers are striking. A typical human-led brokerage team handles 20 to 30 deals a month. Harper handles more than 1,000. The company has served over 5,000 business customers since launch – daycares, manufacturers, car dealerships, local bars and restaurants – the segment traditionally underserved by national brokers because individual accounts are too small to justify dedicated coverage attention.

02 /Why It Matters

YC's argument about insurance agencies is blunt: the labor model is the only reason the economics aren't spectacular. Commission structures haven't changed. What AI does is remove the headcount ceiling. A brokerage that processes 50x the transaction volume at the same labor cost earns 50x the revenue – and in a market as fragmented as commercial insurance, that arithmetic changes the competitive landscape in ways that are currently difficult for incumbents to respond to.

The carrier dimension is less discussed but significant. AI-assembled submissions arrive more complete and better formatted than human-assembled ones. Underwriters who receive consistently clean submissions process them faster – which means Harper's value proposition extends upstream to the carrier. The 160+ carrier relationships Harper has built are simultaneously a product advantage for customers and an efficiency gain for the carrier intake workflow.

Competitors Gyde, FurtherAI, and Vantel are applying AI to the same problem. The differentiation doesn't live in the technology; it builds through carrier relationships, operational depth, and the quality improvements that come from processing thousands of varied commercial accounts without degradation. At 1,000 accounts per month, Harper is likely building that operational muscle faster than any competitor that processes less volume.

03 /Opportunities

Harper is accumulating a transaction data asset that almost no insurance company has: granular, verified SMB commercial risk data at scale. Which coverage types specific business categories actually buy. Where coverage lapses. Which carriers perform in which lines for which geographies. That dataset is worth more as intelligence than as commission revenue.

The path is data monetization alongside distribution. Parametric coverage products designed specifically for the SMB categories Harper knows best – coverage that triggers on measurable events rather than manual claims processes – would be differentiated from anything a general carrier offers. Risk intelligence sold to carriers as a supplement to their own underwriting models is a separate revenue stream. Hippo built the home insurance version of this on weather and geospatial data. Harper is building the commercial SMB version on transaction and risk data at a resolution that doesn't exist anywhere else.