The Growth Podcast

Today’s episode
LinkedIn just changed the title of its product managers to product builders.
What does it even mean to be a “builder PM”?
Well, tools only get you so far. Learning Claude Code is helpful, but means nothing if you don’t have an understanding of the underlying first principles.
That’s today’s episode.
Mahesh Yadav created one of our most popular episodes, with over 35K views on YouTube, and now he’s back. Earlier, he taught you AI agents. Today, he’s touching you how to become a builder PM:
If you want access to my AI tool stack - Dovetail, Arize, Linear, Descript, Reforge Build, DeepSky, Relay.app, Magic Patterns, Speechify, and Mobbin - grab Aakash’s bundle.
I’m giving a free talk on how to get interviews at the top AI PM companies on Thursday, April 23rd 2026 @ 9:00AM PDT. Grab your seat.
----
Check out the conversation on Apple, Spotify, and YouTube.
Brought to you by:
* Maven - Build cohort-based courses that scale
* Amplitude - The market leader in product analytics
* Jira Product Discovery - Prioritize what matters with confidence
* NayaOne - Airgapped cloud-agnostic sandbox to validate AI tools faster
* Product Faculty - Get $550 off their #1 AI PM Certification with my link
----
Key Takeaways:
1. Builder PM defined - A builder PM talks to customers, figures out what to build, and ships the first version to 10 customers without talking to any developer. The skill is knowing what to build, not knowing how to code.
2. Four agent components - Every agent that works has intelligence (model), tools (actions), memory (session context), and knowledge (your company data). Every agent that disappoints is missing at least one.
3. n8n for foundations - n8n is the best learning tool because you visually see every component of the agent architecture as separate nodes. Build your first multi-agent system and evaluation pipeline here.
4. Claude Code ate three company types - Context companies, action companies, and evaluation companies all got replaced by one agentic loop inside Claude Code. The three pieces collapsed into one tool.
5. Computer control is the real unlock - File system access plus bash commands equals full laptop capability. This is why Claude Code went from coding tool to work operating system.
6. Long-horizon jobs changed the game - AI agents went from 3-minute tasks to 3-6 hour sustained jobs in six months. This turns Claude Code from assistant to autonomous worker.
7. Continuous learning loops - Build a second agent that watches your corrections to the first agent's work. After five repeated patterns, it proposes a skill update. Your tools get better every day.
8. OpenClaw pattern - Delegation through existing channels, full machine sandboxing, model-agnostic. Not a product but a pattern that Google and AWS will copy inside their ecosystems.
9. AI PM interviews changed - At L5 and L6, product sense questions are being replaced with live building exercises and system design for AI architectures. Pull out Claude Code during the interview or you are already out.
10. Compensation trajectory - From $120K at Microsoft to $1.3M at Google over 13 years, doubling every 18 months through AI-focused switches. Left because big companies kill innovation with six-week approval cycles.
----
Where to find Mahesh Yadav
* LinkedIn
* Maven Course
Related content
Podcasts:
* Claude Code Team OS with Carl Vellotti
* OpenClaw + Claude Code with Naman Pandey
* Claude Code OS with Dave Killeen
Newsletters:
* The complete context engineering guide
* How to use Claude Code like a pro
* Practical AI agents for PMs
----
PS. Please subscribe on YouTube and follow on Apple & Spotify. It helps!
If you want to advertise, email productgrowthppp at gmail.


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.news.aakashg.com/subscribe

Today’s episode
The design process you learned is already dead.
Most teams still follow the same linear pipeline. Low fidelity to high fidelity to handoff. Sketch it. Spec it. Ship it over the wall. That pipeline was built around a constraint that no longer exists. High fidelity used to be expensive. It is not anymore.
I brought in two people who represent both sides of the new design infrastructure.
Ed Bayes is a member of the design staff at OpenAI. He leads design on Codex, which just crossed 2 million weekly users with usage surging 3X since the start of the year. He spends 70-80% of his time coding. He still calls himself a designer.
Gui Seiz is the Director of Product Design for AI at Figma. He leads design on all their AI features, including the Figma MCP server and Figma Make. His designers are now shipping PRs to production.
----
Check out the conversation on Apple, Spotify, and YouTube.
Brought to you by:
* Bolt: Ship AI-powered products 10x faster
* Amplitude: The market-leader in product analytics
* Pendo: The #1 software experience management platform
* NayaOne: Airgapped cloud-agnostic sandbox
* Product Faculty: Get $550 off their #1 AI PM Certification with my link
----
If you are trying to understand the new design workflow, this is the one episode to watch.
If you want access to my AI tool stack - Dovetail, Arize, Linear, Descript, Reforge Build, DeepSky, Relay.app, Magic Patterns, Speechify, and Mobbin - grab Aakash’s bundle.
I’m putting on a free webinar on Behavioral and AI PM interviews. Join me.
----
Key Takeaways:
1. Code vs canvas is a false dichotomy - The best designers use both fluidly. Canvas for exploration, collaboration, and pixel-perfect craftsmanship. Code for interactions, responsive testing, and the last mile of polish. The question is what you are trying to learn, not which tool to commit to.
2. High fidelity is no longer expensive - The entire linear design process existed because building something interactive required engineering resources. That constraint is gone. A functional wireframe takes the same time as a paper sketch.
3. The Codex-Figma MCP makes handoff lossless - Import screens from a running React app into Figma with exact pixel values. Border radius, padding, shadows, all one to one. It is not a screenshot. It is a responsive, editable design artifact.
4. The reverse direction works seamlessly - Make changes in Figma, paste a component link into Codex, and it updates your code automatically. No redline spec, no handoff document.
5. Ed spends 70-80% of his time coding and still calls himself a designer - The medium changed but the mandate did not. Designers are still the voice of the user, still upholding craft. The tools expanded, the role stayed.
6. Figma designers are shipping PRs to production - Teams that six months ago were AI curious are now banging down the door. Monetization designers who never wrote code are building technically complex prototypes.
7. "Prototypes, not PRDs" is the emerging norm - PMs at OpenAI bring working prototypes to design reviews. They ship PRs to stress-test ideas before handing off to engineering.
8. You do not need permission to start - Someone from OpenAI's GTM team built an iOS app with zero experience. Download Codex and build something for yourself tonight.
9. Curiosity is the defining skill for this era - Not code proficiency, not design talent. The AI is an infinitely patient tutor. Ask questions. Build understanding alongside output.
10. Total football is the mental model - Every player can play every position. Roles still have natural spikes. But the tool constraints that enforced rigid boundaries are dissolving.
----
Where to find Ed Bayes
* LinkedIn
* OpenAI
* X
Where to find Gui Seiz
* LinkedIn
* Figma
* X
Related content
Podcasts:
* Xinran Ma - Design with AI
* Carl Vellotti - Claude Code PM OS
* Codex PM Guide with Carl Vellotti
Newsletters:
* AI prototyping for PMs
* The PM guide to Bolt
* Codex PM guide
----
PS. Please subscribe on YouTube and follow on Apple & Spotify. It helps!
If you want to advertise, email productgrowthppp at gmail.


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.news.aakashg.com/subscribe

Today’s episode
The way PM teams are trending, one PM is going to support 20 people.
Not just engineers. Designers. Analysts. Strategy partners. GTM. Sales. Support.
You cannot answer everyone’s questions about everything. You cannot be in every Slack thread. You cannot be the bottleneck for context that already exists somewhere in a Google Doc no one can find.
But you can give them a high-context, well-organized repo.
Hannah Stulberg is a PM at DoorDash and a former Google PM. She has spent over 1,500 hours in Claude Code.
She wrote the viral Claude Code for Everything series. Her setup is not a personal productivity system. She has structured her entire team’s context into a shared repo that everyone queries.
Her strategy partner - completely non-technical - puts up pull requests every day. Her engineers query metric definitions without asking the analyst. Her designers pull product context without waiting on a PM.
If you are building a team that runs on AI, this is the episode to watch.
----
Check out the conversation on Apple, Spotify, and YouTube.
Brought to you by:
* Bolt: Ship AI-powered products 10x faster
* Jira Product Discovery: Plan with purpose, ship with confidence
* Kameleoon: Leading AI experimentation platform
* Amplitude: The market-leader in product analytics
* Product Faculty: Get $550 off their #1 AI PM Certification with my link
----
If you want access to my AI tool stack - Dovetail, Arize, Linear, Descript, Reforge Build, DeepSky, Relay.app, Magic Patterns, Speechify, and Mobbin - grab Aakash’s bundle.
I’m putting on a free webinar on Behavioral and AI PM interviews. Join me.
----
1. Build a Team OS, not a personal OS - A shared repo where every function checks in work. Engineers, designers, and analysts self-serve without asking the PM.
2. Root CLAUDE.md is everything - Doc index, team roster with Slack IDs, channel map. Keep under one page or you burn context every session.
3. Nested indexes save 97% of context - Every folder gets a navigation CLAUDE.md. A customer query used only 3% of the context window.
4. Three token tiers - Always-loaded root (~500 tokens), folder indexes on navigation (200-500), content files on demand (1,000-10,000+).
5. Split analytics by product area - Metrics, queries, schemas separated. Progressive loading prevents waste.
6. Gate launches on repo updates - Feature not shipped until metrics, queries, schemas, and playbooks are checked in.
7. Verified playbooks kill hallucinations - Analyst-audited methodology. Claude follows verified steps instead of inventing its own.
8. Plan mode makes 10x docs - Shift+Tab twice. Five phases: load context, ask questions, build plan, push thinking, review agents.
9. Split long docs across parallel agents - Each writes to a temp file. Orchestrating agent compiles. Prevents context overflow.
10. The flywheel compounds daily - Automate one task, free time, improve the repo. After 1,500 hours still iterating every day.
----
Where to find Hannah Stulberg
* LinkedIn
* In the Weeds Substack
Related content
Podcasts:
* My Claude Code PM OS with Dave Killeen
* Claude Code + Analytics with Frank Lee
* Claude Code as PM OS with Carl Vellotti
Newsletters:
* The ultimate guide to context engineering
* Build your PM operating system
* How to use Claude Code like a pro
----
PS. Please subscribe on YouTube and follow on Apple & Spotify. It helps!
If you want to advertise, email productgrowthppp at gmail.


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.news.aakashg.com/subscribe

Today’s episode
Claude Code just hit $2.5 billion in annualized revenue in 9 months.
It is the fastest B2B software product ramp in history.
So why are most people still using it like a chatbot?
This is how most people use Claude Code. Type a prompt and get output. The context fills up. It compacts. You lose everything. You start over.
The top users flipped it. They built skills that interview through a framework before building anything. They use sub-agents that preserve context. They have operating systems where every file, every person, every project has a home.
That shift is what today’s episode is about.
I sat down with Carl Vellotti for the third time. His first episode was the beginner course. His second episode was the advanced masterclass. Together they crossed over a million views across platforms.
Today is the operating system layer. If you are already an 80 out of 100 on Claude Code, this episode will bring you to a 95 out of 100.
This episode covers context management, creating sub-agents to manage your context for you, auto-triggering skills with hooks, trustworthy data analysis with Jupyter notebooks, and building an operating system around it all.
If you are living in Claude Code 8 to 10 hours a day and want to stop fighting the tool, this is the one episode to watch.
----
Check out the conversation on Apple, Spotify, and YouTube.
Brought to you by:
* Bolt: Ship AI-powered products 10x faster
* Amplitude: The market-leader in product analytics
* Pendo: The #1 software experience management platform
* NayaOne: Airgapped cloud-agnostic sandbox
* Product Faculty: Get $550 off their #1 AI PM Certification with my link
----
If you want access to my AI tool stack - Dovetail, Arize, Linear, Descript, Reforge Build, DeepSky, Relay.app, Magic Patterns, Speechify, and Mobbin - grab Aakash’s bundle.
I’m putting on a free webinar on Behavioral and AI PM interviews. Join me.
----
Key Takeaways:
1. Context management is the real skill - A single web search eats 10% of your context. Run /context to see what is consuming it. System prompt and MCPs take 10-16% before you type one message.
2. Sub-agents save 20x context - Delegate research to a sub-agent. Same task costs 0.5% instead of 10%. Your main session only gets the summary.
3. Replace MCPs with CLIs - MCPs eat context by existing. CLIs have zero overhead. GitHub CLI, Vercel CLI, Google Workspace CLI are all dramatically more efficient.
4. Powerful skills need zero code - Anthropic's front-end design plugin is just a good prompt. No APIs or tooling. Just rules that tell Claude "do not look like AI."
5. Give Claude self-checking tools - The make slides skill uses Puppeteer to screenshot output, measure overflow, and fix issues before you see them.
6. Repeat prompts for better quality - A Google paper showed pasting a prompt twice helps. Tell Claude to double-check against skill instructions after the first pass.
7. Use hooks to auto-invoke skills - A user_prompt_submit hook matches your words against skill keywords instantly. Zero context cost.
8. Jupyter notebooks solve data trust - Every analysis shows exact code, inputs, and outputs. Traceable and reproducible.
9. Build an operating system - Knowledge folder for people context. Projects folder for task isolation. Tools folder for scripts. CLAUDE.md for identity.
10. The people folder compounds - Connect meeting transcription. After every meeting, update each person's dossier. Every prompt gets more specific over time.
----
Related content
Podcasts:
* Claude Code Masterclass with Carl Vellotti (Ep 2)
* Claude Code PM OS with Dave Killeen
* OpenClaw Setup Guide with Naman Pandey
Newsletters:
* The ultimate guide to context engineering
* How to use Claude Code like a pro
* Claude Cowork and Code setup guide
PS. Please subscribe on YouTube and follow on Apple & Spotify. It helps!


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.news.aakashg.com/subscribe

Today’s episode
Stop applying to AI PM jobs until you understand the fundamentals.
That is not gatekeeping. That is the MIT finding. 19 out of 20 AI pilots fail. The #1 reason? Picking the wrong problem to apply AI to.
Not the wrong model. Not the wrong data. The wrong problem.
Jyothi Nookula has spent 13.5 years in AI. 12 patents. AIPM at Amazon (SageMaker), Meta (PyTorch), Netflix (Developer Platform), and Etsy.
She has hired AIPMs at three of those companies. Trained 1,500+ PMs to transition into AI roles.
If you are trying to break into AI PM, this is the one episode to watch.
----
Brought to you by
* Product Faculty: Get $550 off their #1 AI PM Certification with my link
* Amplitude: The market-leader in product analytics
* Pendo: The #1 software experience management platform
* NayaOne: Airgapped cloud-agnostic sandbox for AI validation
* Kameleoon: Prompt-based experimentation for product teams
----
If you want access to my AI tool stack - Dovetail, Arize, Linear, Descript, Reforge Build, DeepSky, Relay.app, Magic Patterns, Speechify, and Mobbin - grab Aakash’s bundle.
If you want my PM Operating System in Claude Code, click here.
----
Key Takeaways:
1. Two types of AIPM roles exist - 80% are traditional PM roles with AI features added on, where the core product existed before AI. 20% are AI native roles where the product IS AI and the value proposition is impossible without it. Know which type before you apply.
2. The AI PM stack has three layers - Application PMs own user experience (60% of roles, easiest entry point). Platform PMs build tools for other builders (30%). Infra PMs build foundational systems like vector databases and GPU orchestration (10%).
3. 19 out of 20 AI pilots fail from wrong problem selection - AI makes sense for complex pattern recognition, prediction from historical data, and personalization at scale. If explainability is non-negotiable, rules exist, data is limited, or speed is critical, start with heuristics.
4. Most teams overcomplicate their AI technique choice - If you can put the problem in a spreadsheet with inputs and an output to predict, traditional ML is the answer. Perception problems need deep learning. Natural language reasoning needs Gen AI. These are not competitors, they are tools in your toolkit.
5. AI products are fundamentally probabilistic - The same input can produce different outputs. AIPMs must think in quality distributions and acceptable error rates, not binary success vs failure. Data is a first-class citizen, not a nice-to-have.
6. Agents decide, workflows follow steps - Workflows have predetermined sequences with deterministic outcomes. Agents receive goals and independently decide which tools to use. The live N8N demo showed identical tools producing completely different execution patterns.
7. Context engineering is the real production skill - Claude Sonnet has a 200K token context window but that fills fast with knowledge bases, conversation history, and real-time data. Every token costs money. Managing what to load and when directly impacts both quality and cost.
8. Follow the hierarchy before fine tuning - Prompt optimisation first, then context engineering, then RAG. 80% of use cases get solved with RAG. Fine tuning should only be considered after exhausting all three.
9. Build products not projects - Launch your AI work, get real users, encounter real breakage. That gives you richer interview material than any course certificate. Build an agent, build a RAG system, and build an app that solves a real problem.
10. PM culture at big tech shapes who you become - Amazon PMs spend 40-50% of time writing PRFAQs and six-pagers. Meta PMs live in experimentation and statistical significance. Netflix PMs operate with full autonomy through context over control. Each teaches something different.
----
Where to find Jyothi Nookula
* LinkedIn
* NextGen Product Manager
Related content
Podcasts:
* Naman Pandey on OpenClaw
* Lisa Huang on Gemini Gems
* Frank Lee on Amplitude and MCP
Newsletters:
* The ultimate guide to context engineering
* RAG vs fine tuning vs prompt engineering
* AI foundations for PMs
PS. Please subscribe on YouTube and follow on Apple & Spotify. It helps!


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.news.aakashg.com/subscribe