Powered by RND
PodcastsBusinessProduct Growth Podcast

Product Growth Podcast

Aakash Gupta
Product Growth Podcast
Latest episode

Available Episodes

5 of 86
  • The PM’s Role in AI Evals: Step-by-Step
    Today, we’ve got some of our most requested guests yet: Hamel Husain and Shreya Shankar, creators of the world’s best AI Evals cohort.You’ll learn:- Why AI evaluations are the most critical skill for building successful AI products- What common mistakes people are making and how to avoid them- How to effectively "hill climb" towards better AI performanceIf you're building AI features, or aiming to master how AI Eval actually works, this episode is your step-by-step blueprint.----Brought to you by:The AI Evals Course for PMs & Engineers: You get $800 with this linkJira Product Discovery: Plan with purpose, ship with confidenceVanta: Automate compliance, security, and trust with AI (Get $1,000 with my link)AI PM Certification: Get $500 with code AAKASH25----Timestamps:00:00:00- Preview00:02:06 - Three reasons PMs NEED evals.00:04:40 - Why PMs shouldn't view evals as monotonous00:06:23 - Are evals the hardest part of AI products solved?00:07:37 - Why can't you just rely on human "vibe checks"?00:12:11 - Ad 1 (AI Evals Course)00:13:10 - Ad 2 (Jira Product Discovery)00:14:06 - Are LLMs good at 1-5ratings?00:15:45 - The "Whack-a-mole" analogy without evals00:16:26 - Hallucination problem in emails (Apollo story)00:21:22 - How Airbnb used machine learning models?00:23:56 - Evaluating RAG Systems.00:29:52 - Ad 3 (Vanta)00:30:56 - Ad 4 (AIPM Certification on Maven)00:31:42 - Hill Climbing00:35:51 - Red flag: Suspiciously high eval metrics00:39:02 - Design principles for effective evals00:42:42 - How OpenAI approaches evals00:44:39 - Foundation models are trained on "average taste"00:49:36 - Cons of fine-tuning00:51:27 - Prompt engineering vs. RAG vs. Fine-tuning00:53:00 - Introduction of "The Three Gulfs" framework00:56:04 - Roadmap for learning AI evals01:01:41 - Why error analysis is critical for LLMs01:08:29 - Using LLM as a judge01:10:15 - Frameworks for systematic problem-solving in labels01:17:42 - Importance of niche and qualifying clients. (Pro tips)01:18:43 - $800K for first course cohort!01:20:15 - Why end a successful cohort?01:25:49 - GOLD advice for creating a successful course01:33:39 - Outro----Key Takeaways:1. Stop Guessing. Eval Your AI. Your AI isn’t an MVP without robust evaluations. Build in judgment — or you’re just shipping hope. Without evaluation, AI performance is a happy accident.2. Error Analysis = Your Superpower. General metrics won’t save you. You need to understand why your AI messed up. Only then can you fix it — not just wish it worked better.3. 99% Accuracy is a LIE. Suspiciously high metrics usually mean your evaluation setup is broken. Real-world AI is never perfect. If your evals say otherwise, they’re flawed.4. Fine-Tuning is a Trap (Mostly). Fine-tuning is expensive, brittle, and often unnecessary. Start with smarter prompts and RAG. Only fine-tune if you must.5. Your Data’s Wild. Understand It. You can’t eyeball everything. Without structured evaluation, you’ll drown in noise and never find patterns or fixes that matter.6. Models Fail to Generalize. Always. Your AI will break on new data. Don’t blame it. Adapt it. Use RAG, upgrade inputs, and stop expecting out-of-the-box magic.7. OpenAI Doesn’t Get Your Vibe. Their models are average-taste. Your product isn’t. If you want your brand’s voice in your AI, you must define it yourself — with evals.8. Trust LLM Judges... but validate them hard. LLMs can scale your evals — but you still need to verify them against human-labeled data. Don’t blindly trust your judge.9. Your Prompts Are S**T. If your AI is bad, it’s probably your fault. The cheapest, most powerful fix? Sharpen your prompts. Clearer instructions = smarter AI.10. Let AI Teach You. Seriously. LLM judges aren’t just scoring you — they can teach you. Reviewing how your AI fails is the best way to learn what great outputs should look like.----Check it out on Apple, Spotify, or YouTube.----Related Podcasts:Complete Course: AI Product ManagementTutorial of Top 5 AI Prototyping ToolsIf you only have 2 hrs, this is how to become an AI PMCollege Dropout Raised $20M Building AI Tools | Cluely, Roy LeeBolt CEO and Founder on How he Hit $30M ARR in a YearLogRocket CEO and Founder on How to Build a $100M+ AI StartupAmplitude CEO and Founder on Building the Product Analytics Leader----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
    --------  
    1:34:33
  • AI Product Discovery: Complete Course
    We could talk a million things with Tanguy Crusson but I’m keeping it to what I like the most about his work - Product Discovery.He shares a rare, behind-the-scenes look at how his team at Jira Product Discovery uncovers real user problems, validates solutions quickly, and avoids wasting time on ideas that won’t land.----Brought to you by:Jira Product Discovery: Plan with purpose, ship with confidenceProduct Faculty: Get $500 off the AI PM certification with code AAKASH25Vanta: Automate compliance, security, and trust with AI (Get $1,000 with our link)AI PM Certification: Get $500 off Miqdad Jaffer (of OpenAI)’s certification----Timestamps:Preview & Intro — 00:00:00Has AI Actually Changed Product Discovery? — 00:00:27Ad: Jira Product Discovery — 00:10:50Ad: AIPM Certification with OpenAI PM — 00:11:45Where AI Truly Helps in Discovery (Real Use Cases) — 00:12:32Watch the Live Demo on YouTube for Full Context — 00:14:34The Discovery Process (Deep Dive)Phase 1 – Wonder: How Atlassian Uncovers Real Problems — 00:16:51Why Tanguy Hasn’t Written a PRD in 5 Years (+ Wild Engineer Story) — 00:21:04How to Lead Great User Interviews (Tips from a UX Researcher) — 00:25:24Ad: Vanta Compliance & Security — 00:28:30Ad: AI Evals for PMs & Engineers — 00:29:34The Right Way to Structure Discovery Documents — 00:30:34How Atlassian Actually Uses Tools for Product Discovery (Full Stack Demo) — 00:33:23Why This System Works Better Than Traditional PM Workflows — 00:41:24Moving Through DiscoveryPhase 2 – Explore: Rapid Prototypes & Real Feedback — 00:41:58Phase 3 – Make: When the Team Commits to Building — 00:49:04“Just Ship It” is Bad Advice — Here’s Why — 00:52:24Can You Trust Feedback from Free Users? — 00:56:00Phase 4 – Impact: Measuring Real-World Results — 01:07:33How to Build Real Trust as a Product Manager — 01:14:41PMs Want to Keep Engineers Busy: Good Strategy or Trap? — 01:17:46Aakash: “I Wish I Talked to You 10 Years Ago”, Here’s Why — 01:22:46Closing Reflections & Takeaways — 01:26:33----Key Takeaways:1. Discovery isn’t a phase, it’s a system. Atlassian runs product discovery continuously, not just “before development.” It’s embedded across problem finding, prototyping, building, and post-launch.2. Use video, not documents, to communicate user pain. Instead of writing long research summaries, PMs compile 10-minute reels of real customer interviews. Watching raw emotion builds urgency and alignment.3. Start with ~10 users — not thousands. Atlassian validates ideas with small, focused user groups. It's faster, cheaper, and more revealing than wide surveys or launches.4. Prototype with whatever is fastest. From AI tools like V0 to basic Figma slides, the goal is speed. You don’t need polished UIs — you need fast feedback on core concepts.5. Strong user reactions guide investment. When users say “I need this now,” that’s a green light. Mild interest or polite nods? That’s a warning to dig deeper.6. Build only once you have real pull. They don’t move into development (“Make” stage) until a prototype has strong qualitative validation. Code follows conviction.7. PMs rotate weekly to tag and analyze feedback. Every week, one PM owns triaging incoming feedback, tagging it to ideas, and surfacing themes. Discovery is part of the rhythm — not a side project.8. Real discovery requires exposure, not summaries. Dashboards, sanitized reports, and secondhand quotes are not enough. PMs must stay close to raw user input — live or recorded.9. Post-launch reflection is mandatory. Every shipped feature goes into the “Impact” phase. They assess: is it working? Should we scale, refine, or kill it?10. Discovery is a team sport. PMs, designers, engineers, even sales, everyone participates in interviews, watches clips, and shapes the product. It’s not just a PM’s job.----Check it out on Apple, Spotify, or YouTube.----Related Podcasts:How ZoomInfo IPO’d - With CEO/ Founder Henry ShuckHow LogRocket Became a $100M+ AI Company - With CEO/ Founder Matt ArbesfeldHow Amplitude Became the #1 Product Analytics Tool - With CEO/ Founder Spenser SkatesHow Zoom Took Over The World - With Zoom's Meetings Product during Covid ( now leads their Events & Webinars product)----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
    --------  
    1:28:28
  • This $20M AI Founder Is Challenging Elon and Sam Altman | Roy Lee, Cluely
    Amazon banned him. Ivy Leagues kicked him out.And still, he went mega viral, built a $6M ARR AI startup, and raised $15M from a16z… all in a matter of weeks.This might be the craziest founder story you’ll hear all year.If you’re building in a competitive market, struggling to stand out, or just want to learn how to blend controversy with growth, this one’s for you.----Brought to you by:AI PM Certification: Get $500 with code AAKASH25Jira Product Discovery: Plan with purpose, ship with confidenceThe AI Evals Course for PMs & Engineers: You get $800 with this link----Timestamps:21yo Worth $100M+ - 0:00Harvard Kicked Me Out - 3:02Tech Twitter Main Character - 4:33Ads - 6:44Controversial = Views Formula - 8:26Stripper Commercial Brainstorms - 12:47Liquid Glass Before Apple - 16:55User Feedback Drives Product - 19:48Sales Tech & Enterprise - 23:16Cheating in Meetings - 27:19Ad - 28:22How to Fundraise Like Roy - 29:28Brain Chips End Game - 31:05Roy vs Elon vs Sam - 33:12Frat House Culture - 34:44The Cluely Internship - 37:37Are We Getting Dumber? - 38:51Roy Going to Jail? - 40:44Thanks for Watching - 42:18----Key Takeaways:1. Don’t wait for permission. He got kicked out of two top schools and used that energy to build something the world couldn’t ignore. If the system doesn’t reward you, build outside it.2. Design for real behavior, not rules. Interview Coder wasn’t legal or polite, it was effective. It gave users AI help without getting caught. Start with what people actually want.3. Rethink how AI should show up. Stop building chatbots. Build experiences where AI quietly blends into the workflow. Think overlays, not windows. Think invisible, not interruptive.4. Your product doesn’t need to sound safe. “Cheat on everything” wasn’t just a headline, it was a magnet for attention. Don’t fear being bold if it reflects what your product actually does.5. You don’t need a pitch deck if the story tells itself. Cluely raised $15M without running a process. When the traction is undeniable and your product is everywhere, investors come to you.6. Build virality into your operating system. Don’t “hope” something goes viral. Study what’s working. Run daily idea sessions. Create with the expectation that every post could hit 100M views.7. Make your content pass two filters or kill it. Can anyone understand it instantly? Will people feel something strong enough to react? If not, it won’t break through. Keep it simple and emotional.8. Ignore vanity metrics, chase visibility. Don’t waste time measuring click-through rates on content that won’t work next week. Stay focused on showing up everywhere your user lives.9. You don’t need a big team to move fast. Cluely runs on 4 engineers. No designers. No PMs. Still they've $6M ARR. If you can ship fast and learn faster, you’re already ahead.10. Build a product that feels inevitable. This isn’t just for sales teams. Cluely is betting on a world where AI support shows up before you even ask for it. Design for that future now.----Check it out on Apple, Spotify, or YouTube.----Related Podcasts:Bolt CEO and Founder on How he Hit $30M ARR in a YearLogRocket CEO and Founder on How to Build a $100M+ AI StartupAmplitude CEO and Founder on Building the Product Analytics Leader----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
    --------  
    43:06
  • How Zoom Took Over The World
    Zoom grew from $1B to $4.5B in months - but behind the scenes, it was chaos.Learn how they froze features, scaled product teams, and held the platform together under global pressure from the head of product at Zoom meetings.----Brought to you by:Jira Product Discovery: Plan with purpose, ship with confidenceThe AI Evals Course for PMs & Engineers: You get $800 with this linkProduct Faculty: Get $500 off the AI PM certification with code AAKASH25----Timestamps:Preview – 00:00:00How Big Zoom Is – 00:02:12Simplifying Zoom UX – 00:03:40From $1 Billion to $4.5 Billion During Covid - 00:04:38The Three-Month Feature Freeze Story – 00:06:38Aakash Shares His Zoom Fortnite Concert Story – 00:07:30Advice for Product Leaders Facing Hypergrowth – 00:09:25Ad 1 (Jira) – 00:11:29Ad 2 (AI Evals Course) – 00:12:43One Thing He Would’ve Done Differently in His Career – 00:13:25Handling Zoom's Overvaluation, Layoffs, and Stock Price Drops – 00:14:20Future of AI in meetings - 00:16:09Zoom’s Shift into Broadcast and Events – 00:18:55Why Send a Meeting When You Could Record a Loom? – 00:21:35Can Zoom Catch AI Avatar Doing Interview - 00:26:23Ad 3 (AIPM Course) – 00:28:11Using Zoom Like a Pro – 00:28:58Cool Things About the AI Companion – 00:32:27Balancing Simplicity While Shipping Great Features – 00:36:37How PMs Can 10x Their Meeting Management – 00:40:42How to Get Hired at Zoom – 00:44:56What’s Unique About How Zoom Builds Products – 00:51:35Lessons Learned from Running QBRs – 00:54:18Culture at Zoom: Metrics Driven or Not - 00:57:03Zoom Team Adopting Prototyping Tools? – 00:58:53Outro – 01:00:17----Key Takeaways:1. Zoom didn’t just scale, it survived a once-in-a-generation demand shock. In early 2020, Zoom grew from $1B to $4.5B revenue and added 180+ PMs in months. To stay afloat, they froze all feature development for 3 months and focused solely on stability, security, and critical user needs.2. Most teams drown in feature requests. Zoom built a triage system. He created structured buckets to track thousands of bugs, asks, and UI issues. This wasn’t just project management. It was the only way to maintain control during chaos.3. The best PMs don’t just fix, they find the root cause. Instead of reacting to symptoms, he emphasized thinking a layer deeper. What assumption broke? What system failed? Root cause thinking kept Zoom from wasting time on surface-level patches.4. Simplicity beats novelty even in billion-dollar products. Zoom’s team went deep on seemingly minor features like “Raise Hand.” They reworked ordering logic, host controls, and UX friction points, because good meetings are built on small, invisible wins.5. Zoom is no longer just meetings, it’s becoming a media platform. From live webinars to pro events, Zoom is building for marketers, educators, and producers. Tools like Zoom Events, Production Studio, and broadcast integrations are reshaping how large-scale communication happens.6. AI is Zoom’s next act and it’s already driving behavior change. The AI Companion generates instant meeting summaries, action items, and catch-up flows. Internally, teams rely on it heavily — and Beckmann calls it Zoom’s most impactful feature post-COVID.7. Async isn’t the end of meetings, it’s a better tool for the right job. After stepping away from meetings, John personally built Zoom Clips. It’s not here to replace meetings — but for product updates, specs, and visual explanations, async is more efficient and respectful.8. AI avatars aren’t hype, they solve a very real problem. When you're triple-booked with no breaks, avatars offer psychological relief - for both you and the people in the room. Better than being just another muted black screen.9. During layoffs, product leaders become morale managers. In 2023, John had to deliver tough news to his team. His stance: be honest, coach through uncertainty, and remind people that companies aren’t families, they’re teams. Your job is to build trust, not give false comfort.10. Want to stand out in a Zoom interview? Be human, not a highlight reel. Zoom hasn’t seen major AI cheating issues, yet. But John says authenticity matters more than polish. Don’t recite perfect answers. Speak like a real person who’s been in the work, solved real problems, and understands the platform.----Check it out on Apple, Spotify, or YouTube.----Where to Find John:LinkedIn: JohnTwitter: Beckmania----Related Podcasts:How ZoomInfo Grows: The Story, Growth Motion, and Product Strategy with CEO and Founder Henry Shuck15 Steps to Build a $100M+ AI Company (From Someone Who Has Done It) - With the Founder & CEO of AI Startup LogRocket, Matt Arbesfeld----Up NextI hope you enjoyed the last episode with Aman Khan (It was not less than a AI PM Crash Course where we dove into 5 key AI PM skills). Up next, we have episodes with:Dan Olsen - Author, Lean Product PlaybookTanguy Crusson - Head of Product, Jira Product DiscoveryHamel Husain and Shreya Shankar - Hamel Husain: ML Engineer, 20 years in AI; Shreya Shankar: PhD; building DocETLFinally, check out my latest deep dive if you haven’t yet: The Playbook to Land Your First PM Job.----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
    --------  
    1:01:29
  • What I Learned From Filming 80 Podcast Episodes
    Starting this podcast wasn’t an easy decision. It takes up nearly half my time. And for months, it wasn’t making any money.But fast forward to today, and the podcast is now profitable and central to my creator business.So I did something different this week.I recorded a solo episode breaking down everything I’ve learned:- How I built and monetized a niche podcast- The exact growth engine that powers the show today- Some killer AI workflows I’ve learned for your job search and PM job----Brought to you by:Miro: The innovation workspace is your team’s new canvasThe AI Evals Course for PMs & Engineers: You get $800 with this linkLinear: Plan and build products like the bestProduct Faculty: Get $500 off the AI PM certification with code AAKASH25----Timestamps:Part 1: How I Built and Monetized a Niche Podcast0:00 - Exponential Growth Finally Happening1:21 - How I Grew This Podcast4:31 - Learning From 80 Episodes8:46 - How Podcasts Make Money10:50 - My Dream Job: OpenAIPart 2: Breaking Into Product Management & Career Growth13:36 - Roadmap to Becoming a Product Manager14:05 - Why 2025 is Great For PMs21:22 - AI PM Skills Roadmap23:02 - Finding Your First Paying Customer26:12 - Zero to 250K Followers Strategy32:21 - LinkedIn $525K Salary Jump34:24 - High-Effort Content Strategy38:05 - Newsletter Success Stack RankPart 3: Getting PM Jobs & Career Advice40:51 - Presenting Soft Skills on Resume42:46 - Breaking Into US PM Jobs (International Students)46:08 - Is Product Management Right For Me?49:54 - Time Management & Prioritization Mistakes52:15 - MBA to PM Roadmap55:41 - Finding the Right Mentors57:56 - Sprint Prioritization Framework59:39 - Detecting Product Passion vs Money Focus1:01:41 - Platform & AI Product PRDs1:04:09 - Product Management Beliefs That Changed1:05:41 - Newsletter Revenue & Churn Management1:07:02 - Time Management & Focus Strategies1:09:07 - Common Misconceptions About Content Creators----Key Takeaways:01 — Podcast TurnaroundThis podcast is now passing 50K listeners per episode, 11K subscribers on YouTube, and has become central to my creator business. The exponential growth kicked in around episode 80 after consistent investment and improvement.02 — Quality Over EquipmentObsessing over microphones, cameras, and lighting hits diminishing returns fast. The real growth came from improving trailers, clip selection, and backing off endless equipment upgrades to focus on profit generation.03 — Creator Time InvestmentThe podcast now consumes 50% of my time but generates significant monthly profit. Success required treating it as a long-term experiment, reinvesting everything for 50+ episodes before expecting returns.04 — Content Platform StrategyStart with LinkedIn until 10K followers, then add YouTube, then Twitter at 100K, then Medium at 1M. Master one platform completely before spreading to the next for maximum growth efficiency.05 — AI Prototyping RevolutionProduct managers can now build functional prototypes in minutes. Using Gemini 2.5 Pro for PRDs and Bolt.new for development, I built Apple Podcasts video features from strategy to clickable prototype in under 10 minutes.06 — Resume AI MasteryGemini 2.5 Pro dominates resume customization, beating all other models. What used to take 15 minutes of manual customization now takes 80 seconds with better results using proper context and prompting techniques.07 — PM Job Security2025 is an excellent time to enter product management. 75% of PM work involves people coordination and stakeholder alignment that AI cannot replace, making it as secure as doctor/lawyer/engineer careers.08 — Career Switching StrategyUse your domain expertise as an edge — become a product analyst first, then transfer internally, or target companies building tools for your previous industry. Take the level down initially but leverage your customer knowledge.09 — PM Strategy FocusStop being the design-engineering glue person. Your core identity should be product strategy, business metrics, and stakeholder management. Empower teams to answer their own questions while you influence at higher levels.10 — Platform Product PRDsBuilding for developers requires two-step user stories — what developers need from your platform, and what end users need from their products. Include API calls, uptime metrics, and governance considerations in your requirements.----Check it out on Apple, Spotify, or YouTube.----Related Podcasts:Growing to 125K+ on LinkedIn and Cracking B2B Content with Pierre HerubelHow to Win on TikTok and Medium with Professor Aaron Dinin, 100K+ on eachHow I Wrote 3 of My Biggest LinkedIn Posts----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
    --------  
    1:10:25

More Business podcasts

About Product Growth Podcast

The latest insights into how great products grow, how to be a better PM or product leader, and how to get a PM job. www.news.aakashg.com
Podcast website

Listen to Product Growth Podcast, The Property Academy Podcast and many other podcasts from around the world with the radio.net app

Get the free radio.net app

  • Stations and podcasts to bookmark
  • Stream via Wi-Fi or Bluetooth
  • Supports Carplay & Android Auto
  • Many other app features
Social
v7.20.2 | © 2007-2025 radio.de GmbH
Generated: 7/12/2025 - 1:17:56 PM