Podcast Episode Summary

The Growth Podcast with Aakash Gupta

Gemini Gem Masterclass From the Creator Lisa Huang – March 5, 2026

Overview

In this insightful episode, host Aakash Gupta sits down with Lisa Huang, renowned AI Product Manager and creator of Gemini Gems, to dive deep into the practical power of Gemini Gems for product managers, lessons from building major AI features at companies like Apple, Meta, Google, and Xero, and actionable advice for advancing your product management career in the era of AI. Lisa shares behind-the-scenes stories, best practices for maximizing Gemini, tips for building productized agents, and a transparent look at what it takes to land and thrive in top PM roles at leading tech companies.

Key Discussion Points & Insights

1. What Are Gemini Gems? Why Do They Matter?

[00:00, 02:11]

• Lisa Huang: “Gemini Gems are custom versions of Gemini that you can create for your specific use case. ...The key problem that they all have is that they lack context. ... You can think of Gemini Gems as the difference between a general contractor versus a master craftsman.” ([02:11])
• Gems allow users to store context, personalize tone/style, and reduce repetitive prompting – acting as specialized assistants within Gemini rather than a one-size-fits-all LLM.
• Memorable Moment: House-building analogy to explain specialists vs. generalists for LLMs.

2. The Must-Have Gems for Every Product Manager

[00:15, 03:52, 14:08]

• Three Essential Gems:
  1. Writing Clone: “Create a gem that sounds like you... upload your PRDs, emails, and Slack messages so Gemini can help you accelerate your communication tasks.” ([03:52])
  2. Product Strategy Advisor: “...include your company’s strategy docs, market position, competitor analysis – then it can be a little thought partner for you as you’re thinking through different decisions.” ([03:52])
  3. User Research Synthesizer: “What you could do is upload all those user interview transcripts, survey data, and customer support tickets… and just ask the Gemini gem to synthesize all of that for you.” ([03:52])

3. How To Create a Gemini Gem: Step-by-Step

[05:33, 06:04]

• In Gemini, go to “Gems,” click “Create New,” then:
  1. Give clear instructions – specific and detailed prompts yield better results.
  2. Add knowledge – upload context documents and reference files.
  3. Test and iterate – “It may not be perfect the first time... as you use it and iterate it’ll get closer to what you’re looking for.” ([05:33])
• Demo: Lisa walks through creating a Product Strategy Advisor gem and emphasizes personalizing with company documents. ([06:04–07:59])
• Cross-Team Sharing: “You can be that power user in your own company to build it for your teams… share them around.” ([08:36])

4. Behind Gemini Gems: Product Building Process & Differentiation

[09:01, 11:48, 12:14]

• Gemini Gems started before ChatGPT’s Custom GPTs but accelerated after OpenAI’s launch.
• OpenAI focused on an “app store”/monetization paradigm; Gemini focused on personal productivity and team sharing – less about marketplace, more about enabling individuals and teams to maximize the LLM’s utility.

5. Pitfalls & Best Practices for Using Gems

[14:47–15:53]

• Common mistakes:
  • Instructions are too vague – “Not a vague thing like ‘help me write better’... Do spend time on those instructions.”
  • Missing key context files.
  • Not specializing gems enough.
  • Not iterating – “First version is never going to be perfect. … continue to add those tweaks along the way.” ([14:50])
• Persistence: Gems read from uploaded instructions & context; to update behavior, revise those files. ([15:53])

6. Lisa’s Career Journey Across FAANG & Xero: Lessons in PM Growth

[17:14–19:36]

• Lisa’s career guided by curiosity and personal growth rather than a rigid plan:
  “Every step of the way, you’re kind of figuring out the next step... the principle I’ve always used in my career is really curiosity.” ([17:14])
• Finding your own PM archetype and strengths:
  “The important thing is to figure out what are your strengths, what are the things you enjoy about product, where do you want to play in that product space...” ([18:10])
• Key cultures from each company:
  • Apple: Supreme product obsession, high bar for design and detail (“everything bends to the will of having that amazing, amazing product” [19:36])
  • Meta: Experimentation, data, and execution-driven.
  • Google: Most technical, expects PMs to partner deeply with engineering; technical depth a growing PM expectation.

7. Building AI into AR Products: Meta Ray-Ban Story

[23:07–24:59]

• Behind the Feature: Pitched AI Assistant as the “most important feature on these glasses”—not universally agreed upon at first.
• Complexities included executive collaboration, privacy concerns, user interaction design, and engineering limitations (fashion + tech + privacy).
• On-device AI is the Future: More privacy, less latency, feasible as models shrink. ([25:27])

8. PM Skills for the Future: AI, Agents, and Product Judgment

[26:14–38:40]

• “Deeply understand the user need and the tech; find the intersection.”
• Shipping fast, testing, and iterating is vital: “Whatever your assumptions are today, in a month it will change. … Just build, go fast, see what people are doing.” ([26:29])
• Will AI replace PMs? “I pay them for their product judgment... That’s not something I see going to an AI anytime soon.” ([37:39])

9. B2B SaaS & AI Agents: Lessons at Xero

[27:53–36:40]

• JAX is Xero’s “financial super agent”—automates workflows, personalizes to company data, surfaces insights.
• AI must be accurate – especially in finance:
  “Accuracy is not a nice to have... LLMS out of the box are not that great at math, accounting, tax, etc.” ([30:07])
• Blends LLMs, custom code, and human-in-the-loop evaluations to reach production-grade reliability.

10. AI Product Metrics & Agent-Led Growth:

[32:22–36:40]

• Layered measurement:
  1. Baseline quality/Eval (LLM judges, annotators, custom metrics).
  2. Engagement (MAU, retention, CSAT).
  3. Business impact (revenue attribution).
• Agent-led growth is coming but outside Silicon Valley will take time:
  “Before you get to agent-led growth you need companies to actually delegate... right now there’s still a lot of human monitoring.” ([35:26])
• For B2B SaaS leaders: “Everyone is already behind... you should be deeply integrating [AI] agents everywhere.” ([36:54])

Actionable Career Advice for Product Managers

1. Up-skilling & Staying on the Bleeding Edge

[41:37–46:32]

• The PM role is evolving to require builder skills:
  “I think it will become the expectation that PMs are also builders... create the first design, create the first prototype, go build it or code it.” ([40:12])
• Don’t wait for your current job—“Just go use the tools yourself. You can build your own AI products.” ([42:00])
• Tips for standing out:
  • “A lot of people are still talking about it but not really doing it.”
  • Extra diligence stands out—e.g., “I watched three hours of TikTok videos... and summarized key trends for the product.” ([43:25])
• Best way to stay up to date:
  • Use the latest tools, especially in personal projects.
  • Test and apply them in real-world/product contexts.
  • Limit content overload; choose a few high-quality news sources.

2. Breaking Into Top AI PM Roles

[46:32–49:52]

• Lisa’s roadmap:
  • Do the work firsthand, whether as side projects or in current job.
  • Build your network—past collaborators can lead to opportunities.
  • Prepare specifically for FAANG/large company interviews:
    • “Interviews are themselves a skill.”
    • Practice, drill case/product sense/product execution questions.
• “The more you are truly interested... it will show. Do your research.” ([49:59])

Notable Quotes & Timestamps

• “Gemini Gems are custom versions of Gemini that you can create for your specific use case. ...You can think of Gemini Gems as the difference between a general contractor versus a master craftsman.”
  — Lisa Huang ([02:11])

• “First version is never going to be perfect. ...Continue to add those tweaks along the way.”
  — Lisa Huang ([14:50])

• “Every step of the way, you’re kind of figuring out the next step... the principle I’ve always used in my career is really curiosity.”
  — Lisa Huang ([17:14])

• “I pay them for their product judgment... That’s not something I see going to an AI anytime soon.”
  — Lisa Huang ([37:39])

• “You should be deeply integrating agents, AI everywhere because the future of software is not going to look like it does today. And that future is coming very, very fast.”
  — Lisa Huang ([36:54])

• “Just go use the tools yourself. ...You can build your own AI products...There’s no one stopping you.”
  — Lisa Huang ([42:00])

Timestamps for Key Segments

• [00:00] – Introduction to Gemini Gems
• [03:52] – Three must-have Gems for PMs
• [05:33] – How to make a Gem, step-by-step demo
• [09:01] – The origin story and differentiation from Custom GPTs
• [14:47] – Common mistakes with Gems and best practices
• [17:14] – Lisa’s career arc & approach to PM growth
• [19:36] – Comparing product cultures at Apple, Meta, Google
• [23:07] – Building AI into Meta’s Ray-Ban Stories glasses
• [25:27] – On-device vs. cloud AI for AR
• [27:53] – JAX at Xero: AI agents in financial software
• [32:22] – How to measure AI agent success
• [35:26] – Agent-led growth and current state of adoption
• [36:54] – Why all B2B SaaS should integrate AI agents
• [37:39] – Will AI replace product managers?
• [41:37] – Career path and skills for PMs in the AI era
• [46:32] – Breaking into FAANG or leading AI PM roles
• [49:59] – Final career advice

Additional Memorable Moments

• “Treat [your Gems] like an AI mini-product that you’re creating for yourself and iterate; I think is the key.” — Aakash ([16:01])
• Lisa recounts hiring a senior PM who “watched three hours of TikTok videos from small business coaches” as side research ([43:25])
• On interviews: “The more you do it, the more it’ll just feel like second nature and you’ll have seen a breadth of different types of questions.” — Lisa ([49:19])

Conclusion

This masterclass is essential listening for any product manager grappling with the new AI era. Lisa Huang demystifies Gemini Gems, shares concrete frameworks for agent-enabled products, and provides actionable tactics for building a standout PM career—highlighting that adaptability, curiosity, and practical, hands-on fluency with new AI tools are now table stakes for world-class product leadership.