Aakash GuptaThe Claude Code Setup for Non-Technical PMs That Nobody Shows You
At a glance
WHAT IT’S REALLY ABOUT
A four-level Claude Code setup enabling non-technical PMs to ship
- The episode argues non-technical PMs are becoming bureaucratic bottlenecks unless they learn to build and ship alongside engineers using AI tooling.
- Andre presents a 4-level progression: start with Lovable for low-risk prototyping, bridge Lovable to Claude Code via GitHub, move to Vercel + an IDE (often Cursor) for production workflows, then scale with agents, skills, and CLAUDE.md.
- A live demo shows how to bootstrap an app in Lovable, connect it to a GitHub repo, make changes in Claude Code, merge to GitHub, and visually QA updates back in Lovable before publishing.
- To avoid “slop,” the conversation emphasizes two controls: strong infra/security gates on the deployment end and stronger problem-framing rituals (JTBD, Opportunity Solution Trees, MoSCoW) on the discovery end.
- The “Monday morning move” is to get added as a collaborator to a low-risk repo, pick the oldest backlog item, and implement it with Claude Code on a branch to demonstrate leverage without endangering production.
IDEAS WORTH REMEMBERING
5 ideasStart building with a personal, low-risk Lovable project.
Andre recommends beginning with something that won’t harm a company codebase (e.g., a family booking tool) so you can experiment, make mistakes, and learn basic product/engineering concepts without social or technical risk.
Use Lovable as a bridge: prototype visually, then edit code in Claude Code via GitHub.
The workflow is: bootstrap in Lovable → connect Lovable to a GitHub repo → connect Claude Code to the same GitHub account → make changes in Claude Code and merge → Lovable updates for QA, and publishing controls what becomes public.
Don’t get stuck on Git jargon—describe intent and let Claude Code handle mechanics.
For early-stage solo work, Andre suggests telling Claude Code “put this live” or “add this to the codebase,” then using Lovable/Vercel previews to validate outcomes, rather than manually mastering branches and PR flows upfront.
Graduate to Vercel when you need safe, parallel work across multiple branches.
Level 3 is about speed and reliability: Vercel turns each branch into a preview deployment, letting you test changes independently and merge to main only when satisfied, which is harder to manage with Lovable-as-infra.
Cursor can be a comfort layer and a practical troubleshooting tool for non-technical PMs.
Andre likes Cursor’s vertical session/tab UI, while Aakash highlights Cursor’s free agents for debugging setup issues and its smooth GitHub integration, making it easier to manage multiple efforts without feeling lost.
WORDS WORTH SAVING
5 quotesIf you look at a lot of companies that have non-technical PMs in teams, they're bureaucrats. They're stuck in Jira, they're stuck in Linear, they're stuck in PowerPoints. They're not actually building. They're not pushing code. They're not adding features.
— Andre Albuquerque
If you're a non-technical PM and your job is still managing backlogs, your job is still writing issues, your job is still creating decks or PowerPoints to present to someone and you're not really building stuff, honestly, you're being left behind.
— Andre Albuquerque
It's way less scary than just jumping straight into an IDE or jumping straight into, uh, a code environment.
— Andre Albuquerque
The biggest fear a lot of people have is like, oh, vibe coded code is bad code.
— Andre Albuquerque
This is what AI-native teams are doing, is they're working 50% of the time on improving the infrastructure of the machine itself rather than just tweaking feature by feature, right?
— Andre Albuquerque
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome