How Vercel Uses AI w/ VP of Eng Lindsey Simon

Lindsey Simon is the VP of Engineering at Vercel. Its frameworks (Next.js and v0) define how the modern web gets built. Lindsey joined Vercel in 2021, and has seen the eng org grow from 15 to 200+ today.

As AI enters every layer of the stack, Lindsey believes the next frontier for engineering teams isn’t faster code generation (according to him this faster code gen is table stakes). It’s what he calls "vibe-codeable" systems. It’s about how ideas move through code.

“It’s an idea-first world now. AI has gotten to the point where it can oftentimes generate better code than you might have generated by hand - and much more quickly.”

In our conversation we spoke about:

• How AI accelerates his team’s velocity today
• The shift towards personal software
• What he means when he says products need to be more “vibe-codable”
• What Vercel looks for in new engineering hires
• His thoughts on the future of eng orgs

Here are the highlights!

1. Make your systems vibe-codeable

We asked Lindsey to rate AI's impact on engineering velocity within Vercel. He said the difference often comes down to how old a system is.

“Some teams are a five out of five in terms of AI transformation.. some are still a one because we’re a large team with systems that were built five or seven years ago. AI isn’t amazing at those today.”

The fix, according to Lindsey, is to rewrite for flexibility.

“That’s what digital transformation is going to mean for engineering teams over the coming year - make it more easily vibe-codeable, and not just easily, but correctly vibe-codeable.”

When code is tokenized, structured, and refactor-friendly, AI tools like Vercel’s own v0 can generate and improve it at lightning speed. Lindsey thinks the next big unlocks for velocity may not always come from a new model - it will most likely come from a modernized codebase.

2. AI makes ideas the new unit of work

When code generation gets this good, the bottleneck moves upstream.
Execution isn’t the hard part, but design clarity is.

What matters most is the quality of your prompt, and your ability to think in systems. AI has made the “maker-to-manager” gap smaller than ever. If you can articulate what great looks like, AI will help you get there.

Vercel’s internal mantra: "make it work, make it fast, make it good" has taken on new meaning. The new goalpost for “done” isn’t compiling code. It’s crafting something that feels excellent.

“Show us something uniquely that you would do to make this great... that’s what we want to see. We want to see greatness.”

3. Test-driven development just got supercharged

Every engineering generation has its “aha” pattern. For this one, it’s AI-driven test loops.

“The cost of test-driven development has fallen to zero, and the value of test-driven development just went through the roof.”

When AI can instantly write, run, and rewrite code until tests pass, TDD moves from being a mere safeguard to a powerful feedback engine. Lindsey’s rule of thumb: don’t obsess over the implementation. Just make sure it passes the test.

“I don’t care how it works or what it looks like,” he said. “I care that it passes these tests.”

4. Hire for judgment, not keystrokes

AI hasn’t changed who Vercel hires, but it has changed what “good” looks like.

“It hasn’t changed the skills we hire for. There’s still a baseline of being able to write code and produce quality outcomes.”

What’s changed is how engineers use AI to push past the obvious. Candidates are encouraged to use AI tools during interviews. Primarily to showcase taste.

“We tell candidates to be upfront about what they’re using AI for and why. Because if you vibe-coded the entire problem, could you explain how it works to me? That’s what actually matters - that you know how to optimize, how to build a better product.”

5. From personalized to personal software

The popularity of v0 represents a bigger shift in how people build. AI has enabled the "personal software" era.

Lindsey described engineers building tools just for themselves - one-off apps that streamline their own work.

“That’s personal software. I’m one-time vibe-coding this thing to help me get things done. It’s productionized, but it’s mine.”

What this means for builders

Every engineering team will eventually face the same decision: How easily can our systems be vibe-coded? In a world with AI coding agents, whoever shortens the distance between intent and implementation wins.

We had a great time jamming with Lindsey! We’ve been having similar conversations with product and engineering leaders at top engineering orgs (like HeyGen, Wix, Vanta, etc) to unpack how they actually build with AI. You can find previous episodes on YouTube.