Listen to the original at Lenny’s Podcast ↗youtube.com
By Lenny Rachitsky · with Varun Mohan · lennysnewsletter.com · @lennysan on X · YouTube · LinkedIn
Varun Mohan, co-founder and CEO of Windsurf (formerly Codeium), walks through the company's journey from GPU infrastructure to an AI-native IDE used by over a million developers. The conversation covers pivots, lean hiring, enterprise sales, competitive positioning against Cursor, the future of engineering work, and a live demo of the product.
01Key learnings
- Expect your first big bet to be wrong; stay irrationally optimistic about the vision while ruthlessly testing and discarding the underlying assumptions as new information arrives.
- Be willing to pivot fully and quickly when hypotheses break, as Varun's team did when model architectures converged on transformers and infrastructure stopped being a defensible layer.
- Build the interface where AI-generated code gets reviewed, not just where it gets typed; Windsurf's tripled acceptance rate after porting to a custom IDE showed surface design matters as much as model quality.
- Hire only when a team is genuinely underwater; keeping the company slightly dehydrated forces ruthless prioritization and avoids manufactured work and politics that come with excess headcount.
- Prioritize agency as the key skill for builders: people who take initiative and push ideas to working prototypes will outperform those who only describe ideas.
- Enterprise sales can be valuable early; Varun argues technical founders should not dismiss it, and hiring a sales leader once inbound enterprise demand is clear helps scale it.
- Use your own product data as a moat: preference signals from millions of users, especially on incomplete code, let you train models frontier labs have little exposure to.
- Make bets that cannibalize your current product every 6 to 12 months, because incremental feature work alone will not sustain differentiation in a fast-moving category.
“I want the company to almost be like this dehydrated entity. Every hire is like a little bit of water, and we only go back…”Varun Mohan · Lenny’s Podcast · 00:00:16
“We should be cannibalizing the existing state of our product every six to 12 months.”Varun Mohan · Lenny’s Podcast · 01:07:31
“The best thing to do is just get your hands dirty with all of these products.”Varun Mohan · Lenny’s Podcast · 01:10:51
02Frameworks mentioned
Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.