Skip to content
Aakash GuptaAakash Gupta

He Uses 7 Claude Code Agents to Build Apps with 0 Employees

Gabor Mayer is a PM at Google who runs a 21-agent Claude Code development team. In this episode, he walks through a live demo building a production mobile app from zero to TestFlight - Confluence for specs, JIRA for tickets, Figma for design, and Claude Code for development. Full Writeup: https://www.news.aakashg.com/p/how-to-build-a-full-ai-dev-team Transcript: [VERIFY - transcript URL] --- Timestamps: 00:00 AI agents can now run a startup workflow 01:23 Subscribe and AI tools bundle 01:55 Claude Code as your designer, developer, and systems analyst 02:43 Gabor’s 21-agent startup team inside Claude Code 04:57 Inside the system analyst agent 05:52 Live demo: zero to TestFlight 08:42 Prompting Claude to define a good system analyst 10:02 Ads 11:53 Building the system analyst workflow 12:24 Why documentation matters: Confluence, Jira, and MCPs 15:38 Why classic PM skills make you a better AI builder 19:22 The scaffolding that prevents AI spaghetti code 22:23 Setting up project-specific agents in Claude Code 26:26 Dictating the full product spec for the hockey rules app 32:26 Ads 35:36 Why dictation changes the quality of AI specs 47:38 Creating the visual direction in Figma Make 56:00 The idea-to-prompt-to-design-to-app workflow 1:06:21 Claude Code starts building the Figma screens 1:24:06 Frontend epics and Figma-linked tickets appear in Jira 1:48:57 The hockey rules AI app is live 1:53:56 Full recap: Claude, Confluence, Figma, Jira, Simulator, TestFlight 2:03:20 Should PMs get AI PM certificates? 2:08:21 How to create a PM portfolio that helps you land top jobs 2:13:32 How to get started building with AI agents --- 🏆 Thanks to our sponsors: 1. Maven Custom: Go from PM to AI builder with Cloud Code - https://maven.com/gabor/productbuilder 2. Amplitude: The market-leader in product analytics - https://amplitude.com/session-replay?utm_campaign=session-replay-launch-2025&utm_source=linkedin&utm_medium=organic-social&utm_content=productgrowthpodcast 3. Testkube: Leading test orchestration platform - http://testkube.io/ 4. Land PM Job: 12-week experience to master getting a PM job - https://www.landpmjob.com/ 5. Product Faculty: Get $550 off their #1 AI PM Certification with code AAKASH550C7 - https://maven.com/product-faculty/ai-product-management-certification?promoCode=AAKASH550C7 --- Key Takeaways: 1. One-prompt vibe coding fails because of context compression - When you give one agent one massive specification, the model silently drops details it considers lower priority. Your color palette, edge cases, and security requirements disappear. Break work into smaller scoped tasks with dedicated agents. 2. The system analyst agent is the most important agent in any AI dev team - It asks clarifying questions one at a time, documents decisions in Confluence, and maps dependencies before code is written. Without it, every agent operates on partial context. 3. Dictation produces 5x more specification detail than typing - Use voice tools like Super Whisper to describe your app requirements. Even imperfect dictation captures more nuance than careful typing. The AI handles the interpretation. 4. Reusable agents encode institutional knowledge - Every painful lesson, API workaround, and MCP quirk gets saved in the agent markdown file. The next project starts from a position of strength rather than from zero. 5. Attach screenshots to every front-end development ticket - Without visual references, coding agents default to generic AI aesthetics. The Figma link or screenshot is what ensures your brand design actually shows up in the code. 6. Build a Spaghetti Agent for code quality - A dedicated code maintainability agent checks naming conventions, circular references, and comment quality after every sprint. It catches structural problems a PM would never spot. 7. The coding phase is the fastest part of building - Specification, documentation, design, ticket creation, and team review take longer than the actual code generation. Do not skip the front-end work. 8. Sprint organization with dependency mapping is essential - Use tags as a workaround for Atlassian MCP limitations. Map dependencies between tickets so agents build features in the right order. Without sprints, agents build on top of code that does not exist yet. --- 👨‍💻 Where to find Gabor Mayer: LinkedIn: https://www.linkedin.com/in/mayergabor/ Maven Course: https://maven.com/gabor/productbuilder X: https://x.com/gabor_pm 👨‍💻 Where to find Aakash: Twitter: https://www.x.com/aakashg0 LinkedIn: https://www.linkedin.com/in/aakashgupta/ Newsletter: https://www.news.aakashg.com #claudecode #aipm --- 🧠 About Product Growth: The world's largest podcast focused solely on product + growth, with over 200K+ listeners. 🔔 Subscribe and turn on notifications to get more videos like this.

Aakash GuptahostGabor Mayerguest
Apr 28, 20262h 15mWatch on YouTube ↗

CHAPTERS

  1. 0:00 – 1:23

    AI agents as a full startup workflow (PRDs → design → tickets → code → ship)

    Aakash sets the premise: modern AI agents can run an end-to-end startup workflow, not just help with isolated PM tasks. Gabor previews a multi-agent approach that compresses weeks of cross-functional work into hours.

  2. 1:23 – 1:55

    Tooling plug + show logistics (subscription + AI tools bundle)

    Aakash pauses for channel housekeeping and promotes a bundled set of AI tools offered through his newsletter subscription. This segment is primarily informational and sponsor-like.

  3. 1:55 – 2:43

    Claude Code as designer, developer, and systems analyst

    Aakash frames the episode around using Claude Code as multiple roles a typical startup needs. Gabor is introduced as a Google PM who has been experimenting intensely and building apps with agent teams.

  4. 2:43 – 4:57

    Gabor’s 21-agent “company” inside Claude Code

    Gabor breaks down his agent roster and why each specialized role matters. The system analyst is highlighted as the core agent that turns ideas into structured documentation and actionable tickets.

  5. 4:57 – 8:42

    Inside the system analyst agent definition (what it does and why it’s central)

    Gabor opens an agent markdown file to show how the system analyst is instructed to behave. The emphasis is on ambiguity detection, clarifying questions, dependency tracking, and documentation/ticket generation.

  6. 8:42 – 11:53

    Live demo kickoff: idea → “system analyst” in the Claude consumer app

    To make the workflow accessible, Gabor starts in the standard Claude app (voice-friendly) rather than the CLI. He introduces the sample product: an AI chat app that explains IIHF hockey rules using authoritative sources.

  7. 11:53 – 19:22

    Prompting Claude: what makes a good vs bad system analyst

    Gabor has Claude define the system analyst role and distinguish strong vs weak performance. This becomes the template for how he instructs the agent to behave in the rest of the build process.

  8. 19:22 – 22:23

    Why documentation + Atlassian MCP integration prevents “AI spaghetti code”

    Gabor explains why Confluence/Jira are not bureaucracy—they’re scaffolding for maintainability and repeatability. He connects Atlassian via MCP so agents can write docs and tickets directly, and warns against one-shot vibe coding.

  9. 22:23 – 26:26

    Scaffolding the workflow: controlled access, permissions, and project-specific agents

    Gabor shows how he constrains tools and access (only specific Jira/Confluence spaces) and why. He then transitions into Claude Code, highlighting the difference between roleplay in chat vs actual persistent agents in the coding environment.

  10. 26:26 – 47:38

    Dictating the full product spec for the Rule Ask hockey app

    Gabor delivers a long, detailed voice spec covering stack, data sources, RAG behavior, limits, and security. He argues dictation dramatically improves spec depth and speed compared to typing.

  11. 47:38 – 1:06:21

    Design system creation in Figma Make (inspiration → brand/style guide)

    Gabor uses Figma Make to generate a full design brief (typography, colors, components) from inspiration images. He explains why he uses Figma Make for style direction but relies on Claude Code + MCP for higher-quality implementation in real Figma files.

  12. 1:06:21 – 1:24:06

    Claude Code drives real Figma: screens + clickable prototype via UX flow agent

    With MCP connected, agents create the Figma file structure and build screens directly, then automate prototype wiring (interaction arrows). The result is a small set of high-quality screens with minimal manual work.

  13. 1:24:06 – 1:53:56

    From spec to execution: Jira epics, Figma-linked frontend tickets, and backend setup

    Agents create initial backend setup tickets (Firebase basics, domain, secrets), then generate detailed frontend and backend tickets with Figma links/screenshots to avoid generic AI UI. Gabor tags work into sprint-like batches due to MCP sprint creation limits.

  14. 1:53:56 – 2:13:32

    Build, debug, and ship: simulator demo and TestFlight upload

    After sprint execution, Gabor demonstrates the working iOS app in the simulator, including an “observer mode” that reveals retrieval hits, token usage, and latency. He then uploads the first build to TestFlight and recaps the full toolchain end-to-end.

  15. 2:13:32 – 2:15:25

    Career layer: AI PM certificates, portfolios, and how to start building with agents

    Gabor argues certificates matter less than hands-on capability and demonstrable artifacts. He shares how building small real apps creates portfolio proof, plus practical advice on getting started with whatever AI tool you have and leveling up via structured help if needed.

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