How an Indian immigrant with no industry background raised $25M in insurance AI

Executive overview

Most founders pick markets they know. Aman Gour picked insurance knowing nothing about it — and used that outsider lens to question workflows the industry had stopped questioning.

The insight: industries with unstructured data, labor shortages, and shrinking margins are ripe for AI. Picking a vertical early, going deep, and selling in person drove Further AI from zero to $1M ARR and a Andreessen Horowitz Series A.

Outsider ignorance is a first-principles advantage — you're not constrained by "how things have always been done."

Choosing a vertical

• Pivoted during YC away from a "tar pit idea" (AI agents for software testing)
• YC partner Tom Bloomfield pushed them to go vertical and partner with an industry
• Created three criteria for industry selection:
  1. High volume of unstructured data (LLMs can process it faster)
  2. Labor shortages and shrinking margins (drives adoption urgency)
  3. "Why now" on both technology and business dimensions
• Shortlisted mortgage, legal, and insurance — chose insurance
• Drove around the Bay Area with donuts visiting insurance agents to learn workflows firsthand

First-principles product thinking

• Carriers were building web portals for brokers to fill in forms — brokers hated it
• Questioned the assumption: let brokers send emails; have AI fill internal portals instead
• Improved sub-to-quote and quote-to-bind ratios without forcing behavior change
• No industry background meant no assumption that "this is how things are done"
• Insurance is labeled slow to adopt tech, but on the ground people wanted automation

Early sales motion

• Best sales tool: in-person dinner and relationship — people buy you before they buy the product
• "Show me you know me": highly personalised outbound emails, not mass mail
• Reduced friction for buyers: added two-month opt-out clauses on contracts
• Outbound works, but personalisation at scale is what drives response rates

Building the team and company

• Hired for high agency and low drama — attracted people who grind on hard problems
• Shifted personal narrative from "me" to "we" as a second-time founder
• Two goals: build a public company; build a launchpad for everyone who touches Further AI
• Self-described as "Half Monk, Half Machine" — patient and people-first personally, relentless at work