Today’s Episode

The salary for AI PMs is skyrocketing.

Hamza Farooq works with companies like Home Depot, Trip Adviser, and Jack in the Box on their AI strategy. He teaches AI PM courses at Stanford, UCLA, and Maven.

Today, he’s giving you the complete 6-month roadmap to go from no experience to PM at OpenAI or Anthropic.

We built a live AI prototype in 30 minutes (with RAG and agents working). And Hamza breaks down the exact technical skills you need to master.

----

Check out the conversation on Apple, Spotify and YouTube.

Brought to you by:

* Maven

* Amplitude: The market-leader in product analytics

* Vanta: Leading AI compliance platform

* NayaOne

* Kameleoon: Leading AI experimentation platform

----

Key Takeaways:

1. AI PM salaries are skyrocketing - The median total comp for AI PMs is rapidly increasing. But now you need technical depth. Previously, you didn't need to know what RAG is or how fine-tuning works. Now you have to be a jack of all trades.2. We built a working prototype in 30 minutes - Live demo: Lovable for front-end + n8n for workflow automation + RAG connected and working. What used to take days now takes minutes. This is the power of modern AI PM tools.3. Context engineering is more important than prompt engineering - Prompt engineering is what you tell an LLM. Context engineering is how you design the instructions. You combine: system prompt, user prompt, memory (long-term), and RAG. This enables true personalization.4. Know the difference: fine-tuning vs RAG - Fine-tuning = adding new vocabulary (new words). RAG = adding new knowledge (new information). Use RAG for knowledge that changes frequently. Use fine-tuning for vocabulary or specialized response patterns.5. The 5-step architecture you need to master - Step 1: Understand what LLMs are. Step 2: Learn how to build applications. Step 3: Master prompt engineering. Step 4: Implement RAG systems. Step 5: Build agentic systems. Follow this roadmap on repeat.6. Use the three-wave approach for building - Wave 1: Save time (efficiency gains). Wave 2: Better quality (better output). Wave 3: Completely new (novel capabilities). Start with time-savers, progress to quality improvements, end with breakthrough innovations.7. Ask yourself 3 questions before building anything - Does it solve a user problem? Does it solve an organizational problem? Does it align with your business model? If yes to all three, build it. This validates every project.8. Build-first mentality wins - Don't just follow roadmaps. Keep building things. You have to learn by doing. The best way to become an AI PM is to build 10+ projects and see where your products fit in solving real business problems.9. Real-world example: Traversal.ai - Hamza's company works with manufacturers (Amazon suppliers, Jack in the Box, Home Depot). They built an army of agents processing 20,000 SKUs daily with demand forecasts. Results: better inventory optimization, planning, and cost savings.10. Teaching accelerates your own growth - Hamza makes 10-15% of revenue from Maven courses. Why keep teaching? "I teach because I grow." His foundation course builds empathy with users. His developer course uplifts his technical skills by working on real problems with senior engineers.

----

Where to Find Hamza Farooq

* LinkedIn

* Newsletter

Related Content

Podcasts:

* Google AI PM Director drops an AI PM Masterclass

* If you only have 2 hrs, this is how to become an AI PM

* Complete Course: AI Product Management

Newsletters:

* How to Become an AI Product Manager with No Experience

* How to Write a Killer AI Product Manager Resume

* How to become an AI Product Manager

----

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

Every PM needs to master AI prototyping in 2025.

But which tool should you use? And how do you actually prototype effectively?

Alex Danilowicz built Magic Patterns to $1M in revenue in 6 months. Today, we’re putting his tool against the competition live.

We built the same prototype in 5 different tools and graded each one. Then Alex shared the exact workflow his customers use.

----

Check out the conversation on Apple, Spotify and YouTube.

Brought to you by:

Vanta: Leading AI compliance platform

Testkube: Leading test orchestration platform

Kameleoon: Leading AI experimentation platform

Jira Product Discovery: Plan with purpose, ship with confidence

The AI PM Certificate: Get $550 off with ‘AAKASH550C7’

----

Key Takeaways

1. Different tools for different jobs - Magic Patterns excels at visual prototyping, user research, and design system integration. V0/Replit/Bolt excel at full-stack functionality, real APIs, and backend. We tested 5 tools live—V0 won (3.7 GPA), Magic Patterns second (3.6 GPA).2. Define your end goal before opening any tool - Sharing with customers = need design system. Internal validation = skip brand context. Alex's mistake in our face-off? He jumped into building without setting up his preset and wasted time retrofitting ChatGPT's Agent Kit styling later.3. Set up your design system in 5 minutes - Magic Patterns Chrome extension grabs components from Storybook, production sites, or Figma. Click "Convert to Component" and it's available in every prompt. Converts HTML to Tailwind automatically. 5 minutes upfront saves hours later.4. Gather context before prompting - Don't start with blank prompts. Common sources: Jira tickets, PRDs, competitor screenshots, customer feedback. Power users use ChatGPT/Claude to write their Magic Patterns prompts first.5. Use select mode for iterations - Vague prompts waste time. Bad: "Make it better." Good: "Move toast to top-left and make it green." Always click the exact element you want to change. The AI can't read your mind.6. The new product development workflow - Old: Write PRD → Align stakeholders → Build → Pray. New: Build prototype (30 min) → Share link → Test with customers → Iterate → Write PRD with learnings → Build validated solution. Cuts 15+ meetings down to 1.7. AI prototyping cuts failure rates in half - 80% of features don't hit their metrics. You're building blind. With prototypes, you validate: usability, viability, value, drop-offs, corner cases. Before: only test biggest features. Now: test every feature.8. Break out of doom loops - Pattern to avoid: "Doesn't work" repeated 10 times. Repeating the same prompt makes it worse. Use Magic Patterns' /debug command or restart with clearer prompt. Read the AI's output—it's having a conversation.9. Master the 4-step workflow - Step 0: Define end goal. Step 1: Set up design system (if needed). Step 2: Gather context (PRDs, screenshots). Step 3: Iterate specifically with select mode. This workflow helped Magic Patterns hit $1M revenue in 6 months.10. Know when to use each tool - Magic Patterns finished first in speed with best iteration quality. Replit prompted for OpenAI key (more functionality). Use Magic Patterns for: user validation, testing interactions. Use V0/Replit for: backend, real APIs, deployable prototypes.

----

Where to Find Alex Danilowicz

* LinkedIn

* Twitter/X

* Website

----

Related Content

Podcasts:

* Cursor Tutorial

* Windsurf Tutorial

* AI Prototyping Tutorial

Newsletters:

* AI Agents: The Ultimate Guide for PMs

* Ultimate Guide to AI Prototyping Tools

* How to Land a $300K+ AI Product Manager Job

----

P.S. More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

----

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

Why do some AI products feel like magic while others feel like work?

You shipped. It works. Your metrics show “success.”

But users aren’t coming back. They’re not telling friends. And next quarter, they’ll switch to the competitor with a better model.

Nesrine Changuel built Spotify Wrapped and ran Google’s Delight Team. Today, she’s giving you the complete playbook:

The 4-step Delight Model to engineer emotional connection (not just satisfaction)

----

Check out the conversation on Apple, Spotify and YouTube.

Brought to you by:

* Miro: The AI innovation workspace

* Vanta: Leading AI compliance platform

* Testkube: Leading test orchestration platform

* Kameleoon: Leading AI experimentation platform

* The AI PM Certificate: Get $550 off with ‘AAKASH550C7’

-----

Key Takeaways

01 | Understand the 3 Types of Delight

Low Delight solves functional needs only. Surface Delight adds emotion without function (confetti, animations). Deep Delight combines both - solving problems while creating emotional connection. ChatGPT and Cursor win because they nail Deep Delight. Most PMs only ship Low Delight.

02 | Follow the 50-40-10 Rule

Allocate 50% of your roadmap to Low Delight (core functionality), 40% to Deep Delight (differentiation), and 10% to Surface Delight (brand personality). Deep delight drives 2x retention, 2x referrals, and 2x revenue versus satisfied users. This is your competitive moat.

03 | Start with Motivational Segmentation

Stop segmenting by demographics. Identify WHY users actually use your product. Map functional motivators (search, get inspired) AND emotional motivators (feel less lonely, feel proud). Your users aren't all using your product for the same reason.

04 | Use the Delight Grid

Create a grid with functional motivators on vertical axis and emotional on horizontal. Place every feature idea on it. Only functional = Low Delight. Only emotional = Surface Delight. At the intersection = Deep Delight. Can't map it? Don't build it.

05 | Apply the Humanization Technique

Ask: "If my product was a human, how would the experience be better?" Google Meet compared to being in the same room, not Zoom. Dyson compares to hiring a human cleaner, not competitors. This creates features like hand raise and emoji reactions.

06 | Validate with the Delight Checklist

Before shipping, ask: Does it bring value to business AND user? Is it inclusive? Is it familiar? Is it continuous? Is it measurable? Google Meet held back filters until they worked on ALL skin tones. This prevents Apple's breakup message disaster.

07 | Study Deep Delight Examples

Gmail Smart Compose reduces stress while helping you write. Google Meet's AI translation uses YOUR voice and emotion. Spotify's Discover Weekly personalizes while creating belonging. Chrome's Inactive Tabs improves performance while respecting user relationships. Function + emotion together.

08 | Test for Corner Cases Obsessively

Apple's AI summarized a breakup as "no longer in relationship, wants belongings." WhatsApp told a grieving person to "ask John to resend" a photo of her deceased brother. AI progresses fast functionally, but emotional needs get ignored. Corner cases destroy reputations.

09 | Learn from ChatGPT's Win

ChatGPT has 800M users not because of accuracy. People pay subscriptions because they feel less lonely. The emotional need - companionship for solo founders and remote workers - drives retention. Deep delight = personalization that improves over time and remembers context.

10 | Start Delight Early, Not Later

Don't say "let me ship functionality first, add delight later." You're building brand perception from day one. Users forgive functional gaps if the experience delights. They won't forgive boring products that work. Engineer delight from the start.

----

Where to Find Nesrine Changuel

* LinkedIn

* Twitter/X

* Product Delight Book

----

Related Content

Podcasts:

* How to Use Google’s Latest AI Tools

* What it means to be Design-Led

* If you only have 2 hrs, this is how to become an AI PM

Newsletters:

* How to Build AI Products Right

* How to Land a $300K+ AI Product Manager Job

* How to Become an AI Product Manager with No Experience

----

P.S. More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

----

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

Alex Reachvky has helped hundreds of PMs land $700K+ AI jobs.

The gap between $140K and $700K isn’t magic. It’s method.

Today, he breaks down the exact AI-powered workflow to land an AI PM job, from resume creation to acing interviews.

This is the playbook PMs are using right now to 10X their callbacks and land multiple offers.

----

Check out the conversation on Apple, Spotify and YouTube.

Brought to you by:

Brought to you by Linear: Plan and build products like the best.

----

Key Takeaways

1. AI PM jobs pay 30-40% more than regular PM roles - Group PMs make $360K-$600K, CPOs make $2M+. In 2025, 20% of PM roles now mention AI (up from 2% in 2023). The market is exploding and compensation bands are wider than ever.

2. Your resume's top 3 lines are everything - Recruiters spend 7 seconds scanning for Impact, Scope, and Recognizability. Template: "[X years] PM at [Google] | [2B users] | [Scaled revenue 50% YoY to $3.5B]." Pack your biggest wins and recognizable brands here.

3. Create a "bullet vault" then customize in 5 minutes - Use AI to transform your raw career dump into structured bullets covering all PM skill bundles. Use this master resume to tailor for each role by extracting 3-5 non-generic must-haves from the JD.

4. Cold applications get 1% callbacks, outreach gets 10-15% - The math: 30-100 apps → 3-4 callbacks → 1 interview. For every role, find the hiring manager, recruiter, and senior PM. Use ContactOut for emails. Message formula: 1 intro + 3 bullets + 1 CTA under 150 words.

5. Follow up on days 2, 3, and 5 - People miss emails. Persistence wins jobs. If no email response, send LinkedIn connection request. The golden age of PM networking is here: send 30 connection requests daily, comment on posts for 10X more reach than posting.

6. Use Whisper to brain dump at 200 WPM - Answer 24 career questions by speaking instead of typing at 120 WPM. Cover: projects, impact, obstacles overcome, tools introduced, people mentored. This becomes your career vault for both resumes and behavioral interview stories.

7. Behavioral interviews follow Hook-Principles-Action-Results-Learnings - Build 10-15 stories covering leadership challenges, stakeholder conflicts, failed projects, launches. Practice progression: Written first → Spoken → Timed (under 3 minutes). Feed to AI for refinement and probing follow-ups.

8. Case interviews are evaluated on 6 dimensions - Structured Thinking, User Focus, Product Sense, Prioritization, Communication, Creativity. Prompt AI: "You're a FAANG interviewer. Ask me ONE question. Rate 1-5. Quote my weak phrases, explain why they failed, give better approach."

9. Only apply when 50%+ aligned with the role - Extract non-generic must-haves from the JD using AI. Ignore "team player" fluff. Focus on: specific tech infrastructure, growth levers, scale requirements. Rewrite top 3 lines and stack rank bullets to match. Don't let AI fabricate experience.

10. Build your target company list strategically - AI prompt: "Create 50-100 companies ranked by fit. Consider: size (public/late-stage/early-stage), interests, geography." Keep broad. More interviews = better negotiation leverage. Focus on roles posted in last 24 hours. Most PM jobs still in SF Bay/Seattle.

----

Where to Find Alex Rechevskiy

* LinkedIn

* Twitter/ X

* Website

----

Related Content

Podcasts:

* Google AI PM Director drops an AI PM Masterclass

* If you only have 2 hrs, this is how to become an AI PM

* Complete Course: AI Product Management

Newsletters:

* How to Become an AI Product Manager with No Experience

* How to Write a Killer AI Product Manager Resume

* How to become an AI Product Manager

----

PS. More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

----

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

Every PM is asking the same question: Which AI tools actually make me faster?

There are hundreds of apps. Most are hype. Some are game-changers.

Today, I sat down with Anshumanni Rudra - VP of Product at Hotstar, now Group Product Manager at Google leading all APAC payments - to rank 70+ AI tools tier-list style.

We didn’t hold back. S-tier tools got crowned. D-tier tools got exposed.

And we revealed the single best AI tool for product managers in 2025.

Watch the full episode for a chance to win a 1-year free subscription to my newsletter.

----

Check out the conversation on Apple, Spotify and YouTube.

Brought to you by:

* Miro: The innovation workspace

* Vanta: Leading AI security & compliance platform

* Testkube: Leading test orchestration platform

* Kameleoon: Leading AI experimentation platform

* Dovetail: abc

----

Key Takeaways

1. Claude Code is the absolute best AI tool for PMs - Anshumanni runs 6 terminal windows simultaneously doing different things on different parts of his directory. It understands your entire codebase and lets you go from idea to working code in minutes.

2. superwhisper is the S-tier dictation tool that has Anshumanni shouting debugging commands at his screen like Tony Stark. His typing speed has actually fallen since he started dictating everything, making typing feel obsolete.

3. Lindy.AI is the S-tier agent builder PMs actually want because you can prompt it with natural language instead of building flows. Create email responders, meeting prep assistants, and podcast-to-blog converters without touching code.

4. Replit is the ultimate AI prototyping champion that can plan and work for hours building complete applications with minimal guidance. Even before AI, Replit was a strong web-based IDE with deep developer understanding that shows.

5. Granola is the S-tier meeting tool that learns your style and auto-generates talking points based on previous conversations. Unlike Otter or Fireflies, it has intelligent context awareness like a personal assistant.

6. Perplexity gets C-tier as Anshumanni's usage has "gone down quite drastically" since the early days. AI mode in other tools now does what Perplexity used to do with deep search rabbit holes.

7. Cursor gets A-tier as the only IDE with the agent on the right side of the screen, which matters for how PMs think. Anshumanni's usage is "way higher" than other tools purely because of this UX choice.

8. Bolt is Anshumanni's pick for best AI prototyping in A-tier with the best structure from the start. It thinks about both front-end and back-end by default, letting you go from prompt to deployed app in minutes.

9. GitHub Copilot is the first D-tier tool because it's "just not as good" - very ChatGPT focused, not enough Claude. Developers are leaving for Cursor and Claude Code for a reason.

10. Don't chase shiny tools - analyze how you spend your week, find what takes the most time, then find the specific tool that solves that problem. Pick tools for your workflow, experiment, then measure if they improve productivity.

---

Related Content

Podcasts:

* How to Use Google’s Latest AI Tools | Jaclyn Konzelmann Episode

* How to PM Production Changes with Devin | Sahil Lavingia Episode

* Complete Course: AI Product Management

Newsletters:

* How to Become an AI Product Manager with No Experience

* How to Write a Killer AI Product Manager Resume

----

Want my coaching to your dream AI PM job? Apply to grab one of the remaining 17 seats in my cohort:

P.S.1 More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

P.S.2 I’d really appreciate ratings + reviews on podcast platforms as well.

----

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

I had a precious hour of a Google AI PM Director’s time. So, I extracted all the best insights about AI PM for you:

How to use Google’s latest AI tools like an insider

How to build great AI products

How to become an AI PM

And I didn’t hold back on the tough questions. And Jaclyn Konzelmann dropped an absolute masterclass.

You don’t want to miss her advice on AI PM resumes...

----

Check out the conversation on Apple, Spotify and YouTube.

Brought to you by:

Vanta. Pendo. Linear Generic. Jira Product Discovery.

* Vanta: Leading AI security & compliance platform

* Pendo:

* Linear: Plan and build products like the best.

* Jira Product Discovery: Plan with purpose, ship with confidence

* LandPMJob: Land a PM Job with Aakash Gupta

----

Key Takeaways

1. Nano Banana Understands World Models: Ask it to show Toronto in winter → adds snow. San Francisco in winter → no snow. The model knows SF doesn't get snow. This world knowledge unlocks creative workflows beyond basic image generation.

2. The Colorization Workflow: Use Gemini Pro to refine prompts → Focus on vibrant colors, lighting transformation, hyperrealistic detail, modern camera optics → Add negative prompts for failed iterations. "Keep playing around with things until you get it just right."

3. Chain Tools for Advanced Workflows: Photo → Imagen (reimagine as drone show) → Veo (animate the drones flying) → Result: Your pet as a living drone show with tail wagging. Access through AI Studio, Gemini app, or Mixboard.

4. Build AI Apps Without Code Using Opal: Describe what you want in natural language → Opal writes the prompt chains → Customize models and outputs → Share publicly. Examples: Resume critique tool, nature collage generator, custom storybook maker.

5. The Anatomy of an Agent Framework: Every AI agent has 3 components - Models (text/image/video capabilities), Tools (APIs, search, UI actions), Memory (what to remember, personalization strategy). Define these before writing code or PRDs.

6. The User Interaction Spectrum: Every AI product falls on "Do it FOR me" (Deep Research, Audio overviews that run and return) vs "Do it WITH me" (vibe coding, interactive experiences).

7. The Inverted Triangle: Think Big, Ship Fast: Think REALLY big → Use 3 levers to ship: Scope (ruthless MVP cuts), Positioning (beta/experiment labels), Audience (internal → trusted testers → public). Don't let process slow the vision.

8. Ask The Paradigm Shift Question: Are you building a faster horse or a car? Process-improving a workflow or creating an entirely new one? "The real value is the unlock on what's the new way things will get done."

9. The Future-Proofing Question: What happens when models get better? Real example: Mixboard threw out months of image editing work when Nano Banana launched with natural language editing.

10. Google's 6 Hiring Criteria for AI PMs: Exceptional product taste, visionary leadership (think 5 steps ahead), clarity in chaos, compelling product storytelling, full-spectrum execution (blended role profiles), deep AI intuition. Keep resume to 1 page, show actual work, design with personality.

11. The Side Project Strategy: Run 10 side projects simultaneously. Not to launch 10 products, but to think differently and connect dots.

12. Don't Get Precious About Ideas: Any single idea can get commoditized in weeks with AI. The skill isn't having one great idea—it's consistently generating good ideas.

----

Where to Find Jaclyn Konzelmann

* X (Twitter)

* Linkedin

* Substack

----

Related Content

Podcasts:

* How to Become, and Succeed as, an AI PM | The Marily Nika Episode

* If you only have 2 hrs, this is how to become an AI PM

* Complete Course: AI Product Management

Newsletters:

* How to Become an AI Product Manager with No Experience

* How to Write a Killer AI Product Manager Resume

* How to become an AI Product Manager

----

P.S. More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

----

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

Sam Altman said one person will build a billion-dollar company.

Sahil’s already halfway there with just one employee.

Most PMs are still running 6-week sprints. Writing 10-page PRDs. Coordinating between designers and engineers. Sahil ships features from Slack to production in 30 minutes.

Here’s the exact AI workflow powering Gumroad’s $10M ARR:

If AI gets it wrong, your communication was unclear.

----

Check out the conversation on Apple, Spotify, and YouTube.

Brought to you by:

* Vanta: Leading AI security & compliance platform

* Testkube: Leading test orchestration platform

* Kameleoon: Leading AI experimentation platform

* The AI PM Certificate: Get $550 off with ‘AAKASH550C7

----

Key Takeaways

1. Three-Tier AI Workflow: Small tasks (Slack → Devon → Production), Medium tasks (GitHub issue → GPT for PRD → V0 prototype → Ship), Large tasks (4-line brief → V0/Codex → Vercel → Cursor → Production). Match the tool to task complexity.

2. From Slack to Production in Minutes: Customer reports feature request in Slack with screenshots. Type "Devon, address this." Devon reads thread, writes code, opens PR, ships to production. "Weeks of coordination at big companies. We just decide and Devon addresses it."

3. The PRD Is Dying: Stop writing 20-page PRDs for AI. Write 4 lines. Let AI prototype. See what it misunderstands. That reveals what you forgot to specify. "The PRD is only as dense as what cannot be inferred naturally."

4. Use AI to Refine Your Thinking: Paste brief into V0, GPT, and Codex. Each builds something different. Their mistakes show your communication gaps. It's a fake conversation with engineers that makes your real spec better.

5. Architecture Is the New Competitive Advantage: Gumroad is deleting 5,425 lines of CSS to migrate to Tailwind (181 lines). Global CSS means every change affects 300 files. Tailwind means one file change. "Devon made a one-file change. With CSS, you're testing 300 files."

6. Tailwind Is Built for AI: Design system in 181 lines: fonts, colors, padding, borders, shadows. AI never guesses. Industry standard with massive training data. "It's like hiring an engineer who already understands 2x4s. AI knows exactly what to do."

7. AI Is 99th Percentile at Most Things: Defer design and code decisions to AI. If it's important, put it in the spec. If not in the spec, let AI decide. "The decisions AI makes are pretty good. That's why we can move super fast."

8. Work on 5 Things Simultaneously: AI is slow. Solution? Run 4-5 AI sessions at once. While V0 builds, check email. While Codex compiles, answer Slack. "It's like having an army of assistants. I don't wait—I fill the dead time."

9. The Dictatorship Advantage: Big companies need buy-in from PMs, designers, engineers, managers. Gumroad: Sahil → Devon → Production. "The hard part at big companies is aligning people to get behind a decision. It has nothing to do with actually shipping."

10. Perfect the Business, Don't Scale It: $10M ARR, $7-8M EBITDA, $2M dividends last year, 1 employee, 35,000 creators. Goal: $10M EBITDA, then perfect the software. "I just want to work on software, make it better, have people use it, be proud of the work we do."

----

Where to Find Sahil Lavingia

* Linkedin

* X (Twitter)

* Gumroad

----

Related Content

Podcasts:

* We Built an AI Employee in 62 mins

* Conversation with the CEO and Founder of Bolt

* This $20M AI Founder Is Challenging Elon and Sam Altman | Roy Lee, Cluely

Newsletters:

* How to Build AI Products Right

* Ultimate Guide to AI Prototyping Tools

* The Fintech Super App Wars

----

P.S. More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

----

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

Tyler Fisk built a $1.6 million AI education business in one year.

Zero PhDs. Zero Silicon Valley pedigree. Just a systematic approach to building AI agents that actually work in production.

While everyone’s vibe coding in ChatGPT, Tyler’s teaching thousands of students to build multi-agent systems for real businesses. Hundreds of production deployments. Actual revenue.

Today he’s doing a live build: Taking Apple customer service from idea to working multi-agent system in under 90 minutes.

No theory. Pure execution.

----

Brought to you by:

* 1. Maven: Get $135 off Tyler’s course with my code AAKASHxMAVEN

* Vanta: Get $1,000 off AI security & compliance at vanta.com/acos

* Testkube:

* Kameleoon: Leading AI experimentation platform

* The AI Evals Course for PMs: Get $1155 off with code ‘ag-evals’

----

Key takeaways:

1. Stop Vibe Coding: Most teams write one prompt, test twice, ship to production, and hope for the best. Tyler's rule: "We would never put it into production without a human-in-the-loop checkpoint. That's irresponsible." Start with 100% human review, gradually move to 60-70% autonomy.

2. Use Meta-Prompting to Build Agents: Tyler built Gigawatt—an agent with 72,000 characters of system instructions that builds other agents. It researches the domain, writes V1 instructions, evaluates itself (scores out of 100), identifies gaps, and rewrites to V2. Goes from 77% to 86%+ quality.

3. Build Multi-Agent Architectures: Don't build one agent that does everything. Separate concerns like you'd separate teams. For Apple: Core (expert agent, temp=0, finds facts) + Echo (email agent, temp=0.7, writes responses). Each optimized for its specific role.

4. System Instructions Need 7K-9K Tokens: Structure includes Role (job description), Context (business details), Instructions (step-by-step process), Criteria (guardrails), Examples (meta reasoning). Most people write 200 tokens. Tyler writes 7,000-9,000. That's the foundation.

5. Temperature Is Your Secret Weapon: Tyler's Toy Story analogy: Imagine an icy peak in a claw machine. Temp=0 (frozen): claw picks from top only—deterministic, precise. Temp=1 (melted): claw grabs anywhere—creative, varied. Match temperature to agent's job.

6. Information Hierarchy Prevents Hallucinations: Priority order: RAG database first (scraped company docs), System instructions second (built-in expertise), Web search third (with chain-of-verification). When agents search without verification, they hallucinate.

7. Build Complete Workflows: Tyler's 9-step production workflow with 5+ agents: Email arrives → Sentiment analysis (Cinnamon) → Expert research (Core) → Email writing (Echo) → QA loop → Human checkpoint (Slack) → Generative filter → Send → Log to memory.

8. Observational Evals Come First: Test 20+ different scenarios manually. Include edge cases and adversarial inputs. Document every failure. Save golden examples. Only after building confidence do you add systematic evals in production.

9. Calculate ROI as Labor Cost Reduction: Traditional cost: $460/day (expert time + customer service rep + manager review) = $138K/year. AI cost: $153/day (platform fees + API credits + human review) = $45.9K/year. Savings: $92K annual (67% reduction).

10. Emotion Prompting Actually Works: Tyler ends every prompt with "Go get 'em slugger." Based on research: positive reinforcement improves LLM outputs by ~15%. The same psychology that works on humans works on LLMs. "Be nice to your AI. They're gonna have robot bodies soon."

----

Related Content

Podcasts:

* Warp CEO on Profitable AI Agents

* Elizabeth Laraki on AI Product Design

* Claude Code Tutorial

Newsletters:

* AI Agents: The Ultimate Guide for PMs

* How to Build AI Products Right

* Ultimate Guide to AI Prototyping Tools

----

P.S. More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

----

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

Everyone’s building AI products wrong.

They’re sprinkling AI on top like fairy dust. Adding chat interfaces to everything. Ignoring 70 years of design principles.

Elizabeth Laraki was one of 4 designers on Google Search in 2006. One of 2 designers on Google Maps in 2007. She helped create products used by billions—products whose designs barely changed for 15+ years because they nailed it from the start.

Today she breaks down exactly how to design AI features that users actually love.

----

Check out the conversation on Apple, Spotify and YouTube.

Brought to you by:

* Vanta: Automate compliance, manage risk, and prove trust

* Kameleoon: Leading AI experimentation platform

* The AI PM Certificate: Get $550 off with ‘AAKASH550C7’

* The AI Evals Course for PMs: Get $1155 off with code ‘ag-evals’

----

Timestamps:

00:00:00 - Intro

00:01:52 - Elizabeth's background at Google

00:04:19 - Google's AI search integration

00:06:19 - Designing image & video for AI

00:09:44 - AI image expander disaster

00:16:05 - Ads

00:17:50 - AI safeguards & human-in-the-loop

00:18:28 - 3-step AI design process

00:31:29 - Ads

00:33:25 - Designing AI voice interfaces

00:38:25 - Designing beyond chat

00:41:52 - AI design tools for designers

00:44:49 - Live design: LinkedIn for AI

00:57:04 - Google Maps redesign story

01:04:14 - Google Maps India landmarks

01:10:09 - Where to find Elizabeth

01:12:00 - Outro

----

Key Takeaways

1. The Core Design Process Hasn't Changed: Define the product (who, what tasks, what needs), Design it (features, architecture, flows), Build it (UIs, brand). Don't skip to "let's add a chatbot" because you have API access. The fundamentals still apply for AI.

2. AI Adds Non-Deterministic Risk: Traditional software is deterministic - click A, get B every time. AI is non-deterministic with unpredictable outputs. Elizabeth's image expander added a bra strap that wasn't in the original photo. Completely unintentional, completely unacceptable.

3. Work With Research on Safeguards: Audit training data for bias. Build evals that flag sensitive content (human bodies, faces, private information). Show A/B options for ambiguous cases. Make AI's work visible in the UI so users can scrutinize changes.

4. Start With Jobs To Be Done: Don't ask "We have GPT-4, what should we build?" Ask "What painful workflow takes users hours?" Descript mapped video editing lifecycle and baked AI into each job: remove filler words, edit from transcript, create clips, write titles.

5. Map User Context, Not Just Needs: ChatGPT voice in car with three kids? Perfect - nobody's looking at screen. Meta Ray-Bans reading Spanish menu item by item? Terrible - should ask "What are you in the mood for?" Same AI, different context requires different design.

6. Emerge From Ambiguity First: For "LinkedIn for AI," Elizabeth mapped 4 possible directions, picked Matchmaking, identified AI's unlock (personality patterns vs keyword matching), mapped separate UIs for job seekers and employers. Only then touch pixels.

7. Chat Fails for Complex Tasks: Elizabeth tried creating Madrid itinerary in ChatGPT. Every change regenerated everything with new hallucinations. Chat works for Q&A but fails for document creation, visual tasks, multi-step workflows that need persistent editable outputs.

8. Make Chat Supporting, Not Primary: Photoshop embeds AI in existing canvas tools. Google Search shows AI summaries inline in normal results. Cove gives canvas with multiple AI conversations in parallel. Chat is a tool, not THE interface.

9. Stop Adding AI Sprinkles: Elizabeth: "I can't help but think of this massive container of AI sprinkles everybody's shoving on top." Twitter/X + Grok, Amazon + Rufus, Apple Photos all feel forced. Ask three questions: Is this solving a real problem? Does chat make sense? Can you show your work?

10. Google Maps India Innovation: Researched how Indians actually navigate (by landmarks, not street names). Identified which landmarks work (visible from street level like temples, petrol stations). Redesigned entire directions system around that insight. That's design, whether AI or not.

----

Where to Find Elizabeth Laraki

* Linkedin

* X (Twitter)

----

Related Content

Podcasts:

What it means to be Design-Led

Complete Tutorial to AI Prototyping

5 AI Agents Every PM Should Build

Newsletters:

Ultimate Guide to Product Design

Ultimate Guide to AI Prototyping

How to Work With Design for Success

----

P.S. More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

----

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 hit $500 million ARR in four months.

Two product managers. Zero marketing dollars. Just pure viral growth.

While some PMs are still copying and pasting into ChatGPT, others are orchestrating multiple AI agents that work in parallel, automatically reading files, researching competitors, and building prototypes.

Carl Vellotti runs the world’s largest PM Instagram account (55K followers) and has mastered Claude Code better than almost anyone. He’s built his own meme generation system, automated his content workflow, and uses Claude Code for everything from research to prototyping.

Today’s tutorial takes you from beginner to Claude Code hero.

----

Check out the conversation on Apple, Spotify and YouTube.

Brought to you by:

* Linear: Plan and build products like the best

----

Key Takeaways

1/ Stop Working in Chat Windows

Traditional chat requires manually dragging files one at a time. Claude Code lives in your terminal and automatically reads entire folder structures. The interface was the bottleneck all along.

2/ Build Your Knowledge Base First

Create four folders: business-info.md for product context, writing-styles/ for different voices, examples/ for past PRDs, meeting-transcripts/ for automatic uploads. One prompt pulls from everything.

3/ Use the CLAUDE File for Memory

Add rules once, they persist forever. "Never commit without asking." "Always use technical writing." Unlike prompts that get lost in context windows, this stays active every session.

4/ Save Your Best Prompts as Commands

Create /meeting-notes, /competitive-research, /prd-review. Save once, reuse forever. No more hunting through old Twitter bookmarks for that perfect prompt.

5/ Let Claude Plan Before Executing

Press Shift+Tab for Plan Mode. Claude creates full execution plan without touching files. You review, catch mistakes, then approve. This one habit prevents 80% of AI disasters.

6/ Parallelize Everything You Can

Need to analyze 3 customer interviews? Claude spins up 3 UXR agents working simultaneously. Week of manual work becomes 1 hour with parallel agents.

7/ Build Custom Agent Personalities

Designer agent focuses on UX. Engineer agent checks technical constraints. Executive agent evaluates business impact. All three review your PRD simultaneously with specialized perspectives.

8/ Use the $37/Month Combo

Claude Pro ($17) handles research and writing perfectly. Add Cursor ($20) for heavy coding. You get best models for $37 instead of $200/month Claude Max.

9/ Only See Token Usage Here

Claude Code shows real-time token consumption and cost. Finally understand what API pricing actually means. No other interface gives you this visibility.

10/ Start Simple Then Scale

Begin with one research task using file analysis. Add a custom command. Try parallel agents once. You'll never go back to chat interfaces.

----

Where to Find Carl Vellotti

* Linkedin

* X (Twitter)

* Instagram

----

Related Content

Podcasts:

Cursor Tutorial

Windsurf Tutorial

AI Prototyping Tutorial

Newsletters:

AI Agents: The Ultimate Guide for PMs

Ultimate Guide to AI Prototyping Tools

How to Land a $300K+ AI Product Manager Job

----

P.S. More than 85% of you aren’t subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we’ll continue making this content better.

----

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