Lenny's Podcast

The non-technical PM’s guide to building with Cursor | Zevi Arnovitz (Meta)

with Zevi Arnovitz
18 Jan 2026 20 min read 1h 2m
TL;DR

Non-technical PMs can now build serious production apps using Cursor and Claude Code by following a structured workflow: create Linear issues, explore the problem space, plan the implementation, execute with multiple AI models, peer review the code, and document everything as a learning opportunity. The key insight is that code is just words—files on your computer—and the right prompts and workflow beat technical knowledge.

Key Moments
Zevi Arnovitz
“I have zero technical background, did music in high school ... when Sonnet 3.5 came out. I remember watching a YouTube video building apps using Bolt or Lovable. It basically felt like someone came up to me and said, "You have superpowers now."”
Zevi describes the moment he discovered AI-powered coding tools while traveling in Japan
▶ 0:06
Zevi Arnovitz
“You'll be replaced by someone who's better at using AI than you.”
Zevi references the common saying about AI replacing workers, emphasizing skill with tools over credentials
▶ 0:54
Zevi Arnovitz
“It's the best time to be a junior, contrary to what a lot of people are saying, how there's no more junior roles out there. Yeah, that's true, but also when else in history could you get out of school and just build a startup on your own?”
Zevi argues that AI tools have democratized product building for new engineers and non-technical PMs
▶ 0:59
Zevi Arnovitz
“I created a CTO with the custom prompt of it being the complete technical owner of the project. So I told it, "I own the problem. I own how we want the users to feel. You're the complete owner of how this is going to be built. I want you to challenge me."”
Zevi explains how he used a ChatGPT project with a specific system prompt to mimic having a technical co-founder
▶ 10:11
Zevi Arnovitz
“Code is just words at the end of the day. So it's just files on your computer. So basically you can be working on the same project and carry it from app to app.”
Zevi explains why you can graduate between tools like Bolt, Lovable, and Cursor without losing your work
▶ 14:22
About Zevi Arnovitz
›
Zevi is a Product Manager at Meta with zero technical background who taught himself to build serious, revenue-generating products using AI tools like Cursor and Claude Code. He previously worked as a PM at Wix and has created a systematic workflow that non-technical product managers can use to build apps without writing code. His side project StudyMate, an AI-powered study platform, demonstrates that the tools and processes he's developed actually work for real-world applications.
Takeaways
1
**The /command workflow replaces learning to code** Zevi has systematized the entire product development process into reusable prompts (/create issue, /explore, /create plan, /execute, /review, /peer review, /update docs) that run inside Cursor. Each command primes Claude with the exact context and output format needed for that phase. Non-technical PMs can copy these prompts directly and apply them to their own projects without understanding the underlying code.
2
**Cursor + Claude Code beats no-code platforms for serious apps** Bolt and Lovable excel at rapid prototyping because they aggressively write code, but they fail when implementing complex features like payments or database changes. Cursor gives you architectural control—you can split work between multiple models (Composer for speed, Gemini 3 for UI), review code across models, and maintain full project context as your app scales. The tradeoff is exposure to the code itself, which Zevi frames as healthy 'exposure therapy.'
3
**Planning before coding prevents compounding bugs** The exploration phase—where Claude asks clarifying questions about scope, data models, validation, and grading—prevents the gnarly bugs that come from rushing into implementation. This structured approach to discovery mirrors how experienced technical PMs work, proving that good product thinking matters more than code literacy. Zevi spends significant time in the exploration and planning phases before executing any code.