The Idea: Han Wang and Hahnbee Lee met at Cornell in 2020, building a course-planning app that's still used by roughly half the student body today. They were obsessed with the same thing: developers. Not a market, not a category - an audience. By the time they'd graduated, they'd decided to commit full-time to building tools for devs. The problem was figuring out which tool. Their first product, Figstack, was a "Google Translate for code" that hit #1 on Product Hunt - and had retention lower than their morale. So they pivoted. And pivoted again. And kept pivoting. Through seven failed ideas and fourteen months of building with nothing to show for it. At one of their lowest points, a billion-dollar founder told them it had taken his company five years to land their first customer. The advice stuck. On their eighth pivot, they went back to a behaviour they'd seen in Figstack users - pasting AI-generated explanations back into code as docstrings - and asked what a product dedicated entirely to that would look like. This was the beginning of Mintlify.

The Execution:

• 2022: Han and Hahnbee get into YC W22, pivot the day before their interview, and land in the batch with a brand new idea.
• 2022: They spend a weekend building a static-site prototype for documentation. Their first customer, Hyperbeam, pays them after they manually migrate its docs over a weekend. The "do things that don't scale" era begins.
• 2023: Mintlify grows to 1,000+ customers by literally building custom docs for cold prospects before even reaching out. By end of year they're at $1M ARR.
• 2024: They raise an $18M Series A led by a16z in September. Over 20% of every new YC batch starts using Mintlify, creating a viral loop inside the startup ecosystem.
• 2025: Mintlify hits $10M ARR - a 10x jump in 12 months. They acquire Trieve to double down on AI-native search. AI agents now account for 45% of all documentation requests on the platform; Claude Code alone generates nearly 200M requests a month.
• 2026: Mintlify closes a $45M Series B at a $500M valuation, led by a16z and Salesforce Ventures. Over 20,000 companies - including Anthropic, Microsoft, Coinbase, PayPal and Vercel - now run their docs on Mintlify. Content on the platform reaches 100M+ people every year. Han gets a Mintlify tattoo. (No, really.)

The lesson? Pivots aren't failure - they're information. Han and Hahnbee didn't succeed by picking the right idea on day one. They succeeded by picking the right audience (developers) and then being willing to throw out seven ideas in a row until they found the one that stuck. Pair that with a willingness to do embarrassingly manual work (migrating docs by hand, pre-building prototypes for prospects before they'd even said hello), and you get a $500M business on what was originally their 8th attempt.