Aakash Gupta

Sundar Pichai has added $1.11T to Google’s market cap.

His counter-intuitive advice? “You have to reward effort, not outcomes”

It’s basically heresy. “Outcomes, not output” is THE mantra amongst product leaders and books like Inspired. Why is Sundar, a former PM himself, praising efforts over outcomes?

1. Google’s Increased Conservatism
2. What You Can Control
3. The Pendulum Swung Too Far

1. Google’s Increased Conservatism
When Sundar joined Google in 2004, it was very different from when he became CEO in 2015. In 2004, people believed any problem could be solved. And the reward structures were built such that they felt they should try to.

By 2015, Google was not a young unprofitable startup pursuing aspirational bets. It was a conservative profit machine focused on growing its great businesses. This meant folks were not taking the same big, low probability bets Google needed.

The company needed to reward the effort involved with these low probability outcomes.

This made rewarding effort, not output, the solution. The Google advice need not apply in situations where people are risk-taking. But it does in conservative environments.

2. What You Can Control
You can’t control outcomes. You can control outputs. You can control your effort and attitude, learning and quality of features. Outcome focus rewards the lucky and punishes the unlucky.

3. The Pendulum Swung Too Far
Google is famous for its use of OKRs to set goals & evaluate product teams. This approach completely ruled out outputs to focus on outcomes. But it had the result of penalizing good product teams in bad markets. It needed to come to the middle.

In summary, Google is zigging where others are zagging. It’s focusing on outputs and effort, not just outcomes. Consider if you should too.

Everything you need to know about Twitter's algorithm:

🚨 Airbnb just eliminated its traditional product management function.

This is one of the boldest experiments in Product Management we have seen in quite some time.

→ Who’s going to own the roadmap?
→ Who will cross-functional teams collaborate with?
→ Who will be responsible for writing the team’s charter, focus problems, and metrics?

Brian has been posting Twitter threads of each of Airbnb’s two most recent product releases. So it sounds like the executives are re-taking over the roadmap.

And, presumably, engineering managers and designers will split the other tasks. It will be an interesting experiment to watch.

The crowd erupted in applause. Is there something broken with today’s product management? What do you think of the change?

🚨 Breaking: OpenAI just launched a browser.

It's called Atlas. And it's ChatGPT everywhere:

This is the future of AI:

• Fully embedded
• With all the right context
• Where you are, not somewhere else

You can ask it questions to:

1. Help you book travel
2. Pick up where you left off
3. Take actions for you by shopping

Google look out.

Sam Altman and Fidji Simo are coming for you.

Download on MacOS → https://chatgpt.com/atlas

If you set it as your default, they'll boost your limits.

A few resources to help you learn more:

1. Context Engineering → https://lnkd.in/ebfnDUmi
2. Context's Power w Cluely → https://lnkd.in/eEyAzJAF
3. Top Use Cases for PMs → https://lnkd.in/eyfUi8ZN
4. Baking AI into the cake → https://lnkd.in/eDGmsvZ5

Here are 10 super powerful use cases:

1. Instant Competitive Analysis
Open a competitor's pricing page and ask, "Summarize these tiers and compare them to our 'Pro' plan's features from the Notion doc I had open yesterday."

2. User Feedback Synthesis
On a page of App Store or G2 reviews, ask, "What are the top 5 most-mentioned frustrations? Extract 10 quotes related to 'onboarding'."

3. PRD Scaffolding
While viewing a user story in Jira, ask the sidebar, "Draft a list of acceptance criteria & potential edge cases for this feature."

4. Market Research Summarization
Open a 50-page industry report PDF & ask, "What are the top 3 market trends for B2B SaaS in 2026, and which competitors are mentioned?"

5. Technical Elucidation
On a technical blog post about a new API, ask, "Explain this concept like I'm a non-technical stakeholder & list 3 potential use cases for it."

6. "Voice of the Customer" Extraction
With a user interview transcript open, ask, "Extract all direct quotes where the user expresses an unmet need & categorize them by theme."

7. Data-Driven Ideation
Looking at a Amplitude/Tableau dashboard, ask, "This KPI is down 15%. Pull up my recent browser tabs about 'user retention strategies' & summarize the top 5 tactics."

8. Agentic Meeting Prep
"I have a 1:1 with my designer. Open the latest Figma file for the 'New Checkout' flow, the PRD in Notion & the user feedback spreadsheet I was viewing yesterday."

9. Cross-Functional Communication
Highlight a complex engineering update in an email & use the inline assist: "Rewrite this for a marketing & sales audience, focusing on the customer benefit."

10. A/B Test Analysis
On your A/B test results page, ask, "Summarize the results of this test. Is the 'p-value' significant? Based on the-winner, suggest 3 follow-ups from my 'Experiment Ideas' doc."

Follow Aakash Gupta for daily AI PM updates.

The people who jump on this first will get insane productivity advantages.

What's your favorite use case?

The layoff wave tells two stories, not one.

1. Tech giants are cutting to fund GPU purchases.
   
Their revenues are growing. Their stock prices are climbing. They're firing people to free up cash for compute. This isn't cost-cutting during a downturn. It's a forced reallocation from payroll to datacenter capacity. The math is brutal: every percentage point of headcount reduction funds another batch of H100s.

2. Traditional companies are cutting for the opposite reason.

UPS, Nestle, Ford, and Target have already deployed AI tools that work. Customer service automation, supply chain optimization, generative design systems. The productivity gains are real and compounding. These companies don't need to buy massive GPU clusters. They're renting inference from hyperscalers and cutting headcount because the math finally works.

Both sides are feeding the same beast.

Tech companies are buying the shovels. Everyone else is buying the gold those shovels dig up. Semiconductor companies sit in the middle, collecting rent from the entire value chain. TSMC, NVIDIA, and ASML are printing money while employment craters on both ends.

The timing matters.

We're at 10% enterprise AI adoption, heading toward 50%. History says this phase moves fastest and generates the most wealth. But that wealth is concentrating in compute, not labor. The gap between market cap growth and wage growth has never been wider. This isn't a recession. It's a rebalancing. And most workers are on the wrong side of it.

To upskill, you need to master AI:

1. Fundamentals: https://lnkd.in/e6zyYugs
2. Vibe coding: https://lnkd.in/e66DrW-h
3. AI prototyping: https://lnkd.in/eJujDhBV
4. Claude: https://lnkd.in/eZ6UBbAQ
5. AI Agents: https://lnkd.in/eeey5Cxr
6. AI Experimentation: https://lnkd.in/e86mpjGR
7. AI Discovery: https://lnkd.in/e9QrMEDw
8. My coaching: www.landpmjob.com

The path through is in.

In product management, not everything is straight forward maths, or solvable by AI. Yet, some PMs still make better decisions most of the time. How?

That's product sense:

"The ability to find the right solution for the user and business, despite limited and ambiguous information."

I love this definition from Sid Arora.

Give me 5 minutes and I'll explain.

You start by looking at the 10-step PM process:

1. Take a vague & ambiguous problem statement
2. Create, or clarify the overall goal
3. Identify all users in ecosystem
4. Pick 1-2 users
5. Identify major problems of the user
6. Select the problems to solve
7. Brainstorm for solutions
8. Select the highest ROI solution
9. Build and deploy the solution
10. Measure success / collect feedback

Then note product sense is in steps #2-8. It is explicitly NOT about execution, measurement or communication.

It's about 4 tasks:

1. Goal creation

You need to be able to take executive direction and negotiate the goals into something that's realistic given user needs, business demands, and your resources.

2. User discovery

This is the crucial process of figuring out what pulls on the web of stakeholders you're going to make with this product work. And sometimes, it's a learning loop back.

3. Problem discovery

This is the step people like to skip, especially with AI prototyping. But it's important to rigorously understand how users experience their problem today, or iterate.

4. Solution discovery

The step that everyone likes to do: finding solutions that solve the product problem while working for the business. It's what people think is product sense.

So really, product sense is the wisdom to do all 4 of those well, that's developed after years of PMing.

The twist is, this changes for AI.
More: https://lnkd.in/ezbB5CES

📈 Are you looking to improve your product sense?

10 resources for you.

4️⃣ articles:

Product Sense Interview: https://lnkd.in/eTqEFtPM
How to Build AI Right: https://lnkd.in/eDGmsvZ5
How to Build Strategy: https://lnkd.in/e8EUQHAW
Training Product Sense: https://lnkd.in/e8AE5m3B

6️⃣ case studies:

Miro: https://lnkd.in/gVPg-EAq
Figma: https://lnkd.in/eejXM2fp
Airbnb: https://lnkd.in/ef6wKEr2
Klarna: https://lnkd.in/edrTbJUZ
Cursor: https://lnkd.in/eq8xvCAK
Supercell: https://lnkd.in/en2YJBtP

📌 Want my guide to AI product sense? Comment AI product sense + DM me.

What's your favorited definition of product sense?

Miro just previewed the future of product management.

Workshop to AI prototype in minutes:

Traditionally, as PMs we'd run a workshop...
↳ And then wait for designers to iterate on prototypes.

In Miro's new AI Innovation workspace, there's no wait.

The key is how Miro takes the canvas as the prompt. Your canvas is context engineering the prototype spec.

Workshops go from ideation to actual building sessions.

I enjoyed watching the announcement at Canvas 25.

Check it out yourself: https://lnkd.in/da6C2mU7

Emma Craig demoed a single brief generating 3 divergent prototypes. This allows you to more deeply explore the solution space without a prompting step.

This is its power: more solution discovery, less time.

No wonder the company has crossed 100M users.

#𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗺𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁 #𝗖𝗮𝗻𝘃𝗮𝘀𝟮𝟱 #𝗠𝗶𝗿𝗼𝗣𝗮𝗿𝘁𝗻𝗲𝗿

Gemini Head of Product: AI is accelerating product bloat.

Couldn't be more true. Craft has become the key differentiator in product now, not features.

Here's 15 resources to improve your craft:

[1] Frank Chimero on The Web's Grain: https://lnkd.in/eWZajh-d

[2] Addition by subtraction - killing features: https://lnkd.in/e8ZXbjmM

[3] Karri Saarinen's 10 rules for craft: https://lnkd.in/eRqTEd8P

[4] 25 product designs that'll make you jealous: https://lnkd.in/e8YHF_G2

[5] Peter Yang's talk on craft:
https://lnkd.in/exuytZEw

[6] Elizabeth Laraki (ex-Google) on AI craft: https://lnkd.in/eKWCEk8x

[7] Jiro Dreams of Sushi on mastering your craft: https://lnkd.in/ehBHkBys

[8] How to build AI products right: https://lnkd.in/eDGmsvZ5

[9] Paul Graham on taste for makers: https://lnkd.in/e88Dxuwp

[10] How to team with design for craft: https://lnkd.in/e_c8eue5

[11] Brian Chesky on Founder Mode craft: https://lnkd.in/epdjQr22

[12] AI prototyping - vibe code your portfolio: https://lnkd.in/eT3RvQiM

[13] Stripe's philosophy on API design: https://lnkd.in/eAU69FhW

[14] Julie Zhuo on craft in the AI era: https://lnkd.in/eyS8qdQk

[15] Intercom on product craft and principles: https://lnkd.in/eTjU6Yse

📌 Want my free guide to craft? Comment 'craft' + DM.

As Madhu Gurumurthy said, your job as a PM is to cut out the feature bloat and help solve real user problems.

All the more important as we all get hit with AI slop.

Amazon is laying off 30,000 workers this morning.

After recording $60B in profits 💔

These aren't numbers on a spreadsheet.

These are real people - marketers, operators, engineers - entire ladders of people from VPs to new grads.

10% of its white-collar staff.

Suddenly erased.

This is at a time when Amazon is celebrating record profits, and continuing to dominate e-commerce.

The contrast is cruel.

The invites arrive with a sterile calendar invite.

Then, HR reads a script. Access gone. Laptop wiped.

Years of sweat, loyalty, late nights vanishing in a click.

People want to blame AI, but let’s be honest: this isn’t artificial intelligence. This is artificial empathy. The P&L.

If anything, this is Amazon NOT winning in AI enough, getting handily beat by Google and OpenAI.

And this gets to the hard truth: companies will never love you the way you love them.

What matters is your craft, your relevance, your ability to stand tall even when the ground shifts.

My heart aches for everyone being shown the door this morning. The grief is real. The humiliation is real.

But this is a chance to rebuild.

You can take this moment of being treated like “a resource” and turn it into fuel.

Build your craft, create your own leverage.

Route 1: Own your brand

1. Build your LinkedIn: https://lnkd.in/e_2WJTES
2. Create a personal brand: https://lnkd.in/egeXb7Sc
3. Land your dream job: www.landpmjob.com

Route 2: Go startup

1. Learn vibe coding: https://lnkd.in/drV_SxWY
2. Build to production: https://lnkd.in/eNB4mQY7
3. Grow your startup: https://lnkd.in/e9sggnuw

The next time a company pulls the rug, you can stand up, dust off, and say with a straight face: “Their loss.”

And that, maybe, is the truest justice left.

Sad.

Is product sense bullshit? NO! In fact...

It's perhaps the most important skill for PMs:

𝟭. 𝗜𝘀 𝗜𝘁 𝗕𝗦

A Reddit thread asked if product sense is bullshit.

I say, 'absolutely not.'

I fall in the Marty Cagan school of thought:
"It's something you can and need to develop."

𝟮. 𝗪𝗵𝗮𝘁 𝗶𝘁 𝗶𝘀

So if it's not BS, what is it?

Here's how I would define it:

"The skill of consistently crafting products that have the intended impact on users."

𝟯. 𝗛𝗼𝘄 𝘁𝗼 𝗜𝗺𝗽𝗿𝗼𝘃𝗲 𝗬𝗼𝘂𝗿 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗦𝗲𝗻𝘀𝗲

There are 4 major ways:

𝘞𝘢𝘺 1 - 𝘖𝘣𝘴𝘦𝘳𝘷𝘦 𝘱𝘦𝘰𝘱𝘭𝘦 𝘶𝘴𝘪𝘯𝘨 𝘱𝘳𝘰𝘥𝘶𝘤𝘵𝘴 (2-4𝘹 𝘱𝘦𝘳 𝘮𝘰𝘯𝘵𝘩)

• Don't rely on AI unmoderated research sessions alone!
• Actually get out to in-person user sessions
• Even better, modere them yourself

𝘞𝘢𝘺 2 - 𝘋𝘦𝘤𝘰𝘯𝘴𝘵𝘳𝘶𝘤𝘵 𝘦𝘷𝘦𝘳𝘺𝘥𝘢𝘺 𝘱𝘳𝘰𝘥𝘶𝘤𝘵𝘴 (1-2𝘩 𝘱𝘦𝘳 𝘮𝘰𝘯𝘵𝘩)

• Try new products and ask yourself, "WHY!"
• Compare products in the same category
• Reason about features + strategy

𝘞𝘢𝘺 3 - 𝘓𝘦𝘢𝘳𝘯 𝘧𝘳𝘰𝘮 𝘨𝘳𝘦𝘢𝘵 𝘱𝘳𝘰𝘥𝘶𝘤𝘵 𝘵𝘩𝘪𝘯𝘬𝘦𝘳𝘴

• I personally learnt most from following great thinkers
• Especially extraordinary PMs I worked with (and for)
• Thinkers like Julie Zhuo & Marty Cagan also help

𝘞𝘢𝘺 4 - 𝘚𝘵𝘶𝘥𝘺 𝘦𝘮𝘦𝘳𝘨𝘪𝘯𝘨 𝘵𝘦𝘤𝘩 𝘵𝘳𝘦𝘯𝘥𝘴

• Things like AI, cloud, mobile, the internet keep coming
• Stay ahead of them by thinking ahead on products
• Don't just follow the news, reason about the future

A few resources to help you with each:

𝘞𝘢𝘺 1 - 𝘖𝘣𝘴𝘦𝘳𝘷𝘦 𝘱𝘦𝘰𝘱𝘭𝘦 𝘶𝘴𝘪𝘯𝘨 𝘱𝘳𝘰𝘥𝘶𝘤𝘵𝘴

My observation guide: https://lnkd.in/e8AE5m3B
How to interview customers: https://lnkd.in/eAKu8qDp

𝘞𝘢𝘺 2 - 𝘋𝘦𝘤𝘰𝘯𝘴𝘵𝘳𝘶𝘤𝘵 𝘦𝘷𝘦𝘳𝘺𝘥𝘢𝘺 𝘱𝘳𝘰𝘥𝘶𝘤𝘵𝘴

2 examples of the depth to go to

Notion deconstruction: https://lnkd.in/eNYgx74i
Figma deconstruction: https://lnkd.in/eejXM2fp

𝘞𝘢𝘺 3 - 𝘚𝘵𝘶𝘥𝘺 𝘎𝘳𝘦𝘢𝘵 𝘗𝘳𝘰𝘥𝘶𝘤𝘵 𝘛𝘩𝘪𝘯𝘬𝘦𝘳𝘴

Marty Cagan interview: https://lnkd.in/eBRg4tVm
Julie Zhuo interview: https://lnkd.in/eVeEKcip

𝘞𝘢𝘺 4 - 𝘚𝘵𝘶𝘥𝘺 𝘦𝘮𝘦𝘳𝘨𝘪𝘯𝘨 𝘵𝘦𝘤𝘩 𝘵𝘳𝘦𝘯𝘥𝘴

Multi-agent AI: https://lnkd.in/eWebg8gq
Claude Code: https://lnkd.in/eUyPEAma

𝟰. 𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗔𝗜 𝗶𝗺𝗽𝗮𝗰𝘁 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗦𝗲𝗻𝘀𝗲?

Now that the price of a feature is approaching zero at the limit, the most important thing is a product sense.

You must have the ability to prune AI feature slop.

And for AI features, you need a unique version of product sense attuned to its unpredictability.

📌 Comment 'AI Sense' + DM me for my guide to AI product sense.
➕ Follow Aakash Gupta for free daily AI PM knowledge.

I reckon product sense is not just BS, but more people need to actively be cultivating their product sense!

Do you think product sense is BS? Would love to hear 👇

Want to pass PM Interviews at OpenAI, xAI, Meta, Google, Airbnb, Dropbox, Palantir?👇

1. Understand these important concepts:

- Business: P&L, metrics trees
- Technical: APIs, scalability, performance
- Roadmapping: now/next/later, alignment
- AI/ML: prompt eng, RAG, fine-tuning, evals
- Experimentation: when to A/B test, p-values
- Prioritization: RICE, value vs effort, opp cost
- Discovery: customer interviews, user research
- Metrics: north star, guardrail, leading v lagging
- Product strategy: future trends, big tech choices

2. Practice these case interview questions:

- Product sense: https://lnkd.in/eTqEFtPM
- AI product sense: https://lnkd.in/ekWty_bw
- System design: https://lnkd.in/etTmzNzz
- Product design: https://lnkd.in/encHB-k7
- Product execution: https://lnkd.in/e8t8zS9q
- Product estimation: https://lnkd.in/eZvjP4RR
- Product strategy: https://lnkd.in/eqzpahmW
- Success metrics: https://lnkd.in/efdj_QYT
- Vibe coding: https://lnkd.in/e66DrW-h

3. Practice these behavioral questions:

- Tell me about yourself: https://lnkd.in/eYmRYF2Y
- Why want to work here: https://lnkd.in/eeuRsR9M
- What's your current pay: https://lnkd.in/gS-CKH9P
- Improve app speed: https://lnkd.in/dG9EFhiR
- Greatest weakness: https://lnkd.in/ePQ7dBtr
- Were you laid off: https://lnkd.in/e2iTP3xs
- Leadership style: https://lnkd.in/eCPdQVp2
- What level targeting: https://lnkd.in/e2YQ8tK2

4. Study these company specific processes:

- Microsoft: https://lnkd.in/eXWWYefH
- Amazon: https://lnkd.in/eA_WxXHQ
- Apple: https://lnkd.in/eXWWYefH
- Google: https://lnkd.in/ezAHYXhV
- OpenAI: https://lnkd.in/ebR_Auss
- Meta: https://lnkd.in/evHQ_w-E

Image: Yangshun Tay

———

♻ Repost to help others in your network
📕 Save the post so you can find it in future
📌 Comment 'Interview guide' + DM me for all the links
💡 Follow me Aakash Gupta and my company account Product Growth for 2x daily AI PM insights

The way Anthropic builds product is wild.

It's a window into the future of PM:

Catherine Wu, product lead, shared their process:

1. Idea to Prototype

They skip the spec step altogether and build a working prototype using Claude Code.

2. Internal Launch

They ship that prototype internally to everyone. People start dogfooding and giving feedback.

3. Iterate Based on Dog-Fooding

They use the qualitative and quantitative feedback from the internal release to decide what to do next.

This is "prototype first" development.

Loved discovering this from Sachin Rekhi.

Does it work for everyone? No. If you can't get religious dogfooding within the company, or don't trust your AI coding agents, you're going to be set up for failure.

There's a lot of cases where a spec still makes sense.

But, if you can work this way, it's magic.

Interested in making the shift?

8 resources you need:

1. AI Prototyping → https://lnkd.in/eJujDhBV
2. Prototype to Production → https://lnkd.in/eNB4mQY7
3. v0 tutorial → https://lnkd.in/ed6H-wd8
4. Claude Code tutorial → https://lnkd.in/eUyPEAma
5. AI PRDs → https://lnkd.in/eMu59p_z
6. Vibe Experimentation → https://lnkd.in/e86mpjGR
7. AI Customer Intelligence → https://lnkd.in/eKhtZw6s
8. AI Roadmaps → https://lnkd.in/eEuNcj-m

Dream of working at Anthropic? I've helped others.

Let me coach you: www.landpmjob.com

📌 Want my AI prototyping + PRD framework? Comment 'AI PRD' + DM me. Repost to cut the line.

Your best product decisions come from real user behavior. A working prototype is the fastest way there.

That's why this approach is genius.

I applied to 1000+ jobs.

Got rejected from all of them.

I decided to try a new approach, and the result was interviews at Meta, Doordash, and Uber + an offer from Google with a $70k+ raise.

Here's the 9 step system I used to make it happen:

(And how to use AI at every step)

Repost and save before anything else ♻️

𝟭. 𝗗𝗲𝗳𝗶𝗻𝗲 𝘆𝗼𝘂𝗿 𝘃𝗮𝗹𝘂𝗲 𝘁𝗵𝗲𝘀𝗶𝘀

You want to become the obvious bet.

Write a 2-sentence value thesis focused on business impact. End with one metric proof point.

Feed your resume to ChatGPT for help.

🔗 Phyl Terry's masterclass: https://lnkd.in/eVpXWuRN

𝟮. 𝗦𝗰𝗼𝗿𝗲 & 𝘀𝗵𝗼𝗿𝘁𝗹𝗶𝘀𝘁 𝘁𝗮𝗿𝗴𝗲𝘁𝘀

Build a 10-company scoring sheet: Goal, 6-mo Initiative, Challenge, My Angle, Warm Paths.

Use Perplexity to pull goals from earnings calls in seconds.

🔗 Guide: https://lnkd.in/evQeJZGB

𝟯. 𝗕𝘂𝗶𝗹𝗱 𝘆𝗼𝘂𝗿 𝗿𝗲𝗳𝗲𝗿𝗿𝗮𝗹 𝗴𝗿𝗮𝗽𝗵

Map 10-15 warm connections per company: hiring managers, skip-levels, peers, cross-functional partners.

Export LinkedIn contacts to ChatGPT to generate icebreakers.

🔗 Guide: https://lnkd.in/eU8Y_ar9

𝟰. 𝗩𝗮𝗹𝘂𝗲-𝗳𝗶𝗿𝘀𝘁 𝗼𝘂𝘁𝗿𝗲𝗮𝗰𝗵

3-touch sequence:

a. Comment on content →
b. Ask advice question →
c. Offer insights for feedback

Give ChatGPT their signal + your angle for copy.

🔗 Step-by-step: https://lnkd.in/ekvY_gjT

𝟱. 𝗗𝗶𝘀𝗰𝗼𝘃𝗲𝗿𝘆 𝘁𝗵𝗮𝘁 𝗳𝗶𝗻𝗱𝘀 𝗿𝗲𝗮𝗹 𝗽𝗮𝗶𝗻

Ask: What's your 6-month goal? What's blocking it? What's baseline vs. target?

Then use ChatGPT to generate discovery questions that extract metrics & constraints.

🔗 Guide: https://lnkd.in/e8WkHDbK

𝟲. 𝗕𝘂𝗶𝗹𝗱 𝗮 "𝗿𝗲𝗰𝗲𝗶𝗽𝘁𝘀" 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗽𝗮𝗰𝗸

Gather 6 receipts: 2 exec quotes, 1 earnings metric, 1 competitor move, 1 customer pain, 1 industry stat.

Drop the company and problem into Perplexity for sourced receipts.

🔗 How: https://lnkd.in/euH_Q3xR

𝟳. 𝗦𝗵𝗶𝗽 𝗮 𝟲-𝘀𝗹𝗶𝗱𝗲 𝗽𝗶𝘁𝗰𝗵

Structure:

Problem →
Evidence →
2 Ideas →
Plan →
Impact

Ask for feedback, not a job.

Let ChatGPT draft the outline from your research.

🔗 Guide: https://lnkd.in/eEpKEQWM

𝟴. 𝗙𝗼𝗰𝘂𝘀 𝗼𝗻 𝗾𝘂𝗶𝗰𝗸-𝘄𝗶𝗻 𝗽𝗮𝘁𝘁𝗲𝗿𝗻𝘀

Prove value: audit flows, prototype features, analyze competitors.

Hand ChatGPT the goal & constraints for 5 patterns.

🔗 Guide: https://lnkd.in/efbppEDj

𝟵. 𝗖𝗼𝗻𝘃𝗲𝗿𝘁 𝘁𝗼 𝗶𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 𝗺𝗼𝗺𝗲𝗻𝘁𝘂𝗺

Prepare your talk track, STAR stories, and pre-answers to feasibility/data/GTM/ethics/timeline.

Feed them to ChatGPT to improve.

🔗 Lewis Lin's guide: https://lnkd.in/e4MkWiyD

This system works because you demonstrate value before asking for anything.

📌 Comment 'Pitch' + DM for the step by step system.

Want me to coach you on this? Join my cohort: www.landpmjob.com

Follow Aakash Gupta for daily tips

AI product sense is not the same as product sense.

But it's still key to building good products.

If I were looking to learn my AI product sense again, here's the roadmap I would take:

1. What is Product Sense

It's making consistently good product decisions

🔗 Master Product Sense: https://lnkd.in/eTqEFtPM
🔗 Product Strategy Masterclass: https://lnkd.in/ecVF-y6i
🔗 Product Leadership Guide: https://lnkd.in/ewTn7W6z

2. AI Product Sense

AI is different:

• It's probabilistic
• There's cost variables
• And you must build evals

Thus, AI product sense is: the fit between need × model behavior × economics that delivers value

🔗 Become an AI PM: https://lnkd.in/efqv4qUc
🔗 AI Product Strategy: https://lnkd.in/egemMhMF
🔗 How to Build AI products: https://lnkd.in/eDGmsvZ5

It's really different from product sense.

You need to develop 6 lenses:

User Reality
Model Behavior
Economics
System Design
Trust & Liability
Go-to-Market

🔗 Deep Dive: https://lnkd.in/egemMhMF
🔗 AI PRD Template: https://lnkd.in/eMu59p_z
🔗 AI Product Strategy: https://lnkd.in/egemMhMF

Lens 1: User Reality

You need to understand the actual user job

🔗 Jobs to be Done: https://lnkd.in/eD3GzVeq
🔗 Customer Interviews: https://lnkd.in/eAKu8qDp
🔗 Know Your Users: https://lnkd.in/eUuMJZVi

Lens 2: Model Behavior

You have to understand model behavior under variance

🔗 AI Evals Guide: https://lnkd.in/eGbzWMxf
🔗 AI Testing: https://lnkd.in/eKhmKYNp
🔗 LLM Judge: https://lnkd.in/ez3stJRm

Lens 3: Economics

You need to consider the costs at volume.

🔗 Prototype -> Production: https://lnkd.in/eNB4mQY7
🔗 System Design: https://lnkd.in/etTmzNzz

Lens 4: System Design

You must know how to architect to absorb variance

🔗 Prompt engineering: https://lnkd.in/d_qYCBT7
🔗 RAG vs fine-tuning: https://lnkd.in/ebfnDUmi

Lens 5: Trust & Liability

You must understand cost of getting things wrong

🔗 AI Observability: https://lnkd.in/e3eQBdMp
🔗 Building Agents: https://lnkd.in/eeey5Cxr

Lens 6: Go-to-Market

You have to actually get the feature in people's hands

🔗 Product Launch: https://lnkd.in/eB7s6umA
🔗 Go-to-Market Strategy: https://lnkd.in/e9sggnuw

📌 Want my AI product sense framework? Comment 'framework' + DM me.

- - -

I'm coaching 30 people to their dream AI PM job.

Join my cohort: www.landpmjob.com

14 seats left. Starts in 7 days.

- - -

This is one of the most important skills for AI PMs.

If you want to become good at AI product management, check out these 13 resources 👇

1/ Marty Cagan's "AI Product Management"
🔗 https://lnkd.in/eQhETu6g

2/ How to Become an AI PM
🔗 https://lnkd.in/e3akMRm8

3/ a16z's "Getting Started with AI"
🔗 https://a16z.com/ai-canon/

4/ AI Product Strategy
🔗 https://lnkd.in/egemMhMF

5/ Eugene Yan's "Patterns for Building LLM Apps"
🔗 https://lnkd.in/ep8BWuhz

6/ AI Evals
🔗 https://lnkd.in/eGbzWMxf

7/ Chip Huyen's "Building LLM Apps"
🔗 https://lnkd.in/eyNjQdJb

8/ Prompt Engineering Guide
🔗 https://lnkd.in/d_qYCBT7

9/ Hamel Husain's "Your AI Product Needs Evals"
🔗 https://lnkd.in/eJvE8hAS

10/ AI PRD Guide
🔗 https://lnkd.in/eMu59p_z

11/ If This 81 Minute Video Doesn't Make You an AI PM, I'll Delete My Channel
🔗 https://lnkd.in/epkZCNTm

12/ How to Build AI Products
🔗 https://lnkd.in/eDGmsvZ5

13/ Complete Course: AI Product Management
🔗 https://lnkd.in/ezTvSaEX

What resources would you add?

📌 Save the list and share with your network

💼 Ready to land your next PM role? Join my job search cohort, starting Nov 1: www.landpmjob.com

🛠️🤖 How to build AI agents from scratch
(Even if you've never done it before.)
𝗧𝗵𝗲𝘀𝗲 𝗮𝗿𝗲 𝘁𝗵𝗲 𝟴 𝘀𝘁𝗲𝗽𝘀 𝘁𝗼 𝘁𝗮𝗸𝗲, 𝗳𝗿𝗼𝗺 𝗽𝘂𝗿𝗽𝗼𝘀𝗲 𝘁𝗼 𝗨𝗜.

𝙁𝙞𝙧𝙨𝙩, 𝙬𝙝𝙖𝙩 𝙞𝙨 𝙖𝙣 𝘼𝙄 𝙖𝙜𝙚𝙣𝙩?

An AI Agent is an LLM that:

• Reasons on its own
• Act (via calling tools)
• Observes & improve itself

This is the core "ReAct" framework Google released all the way back in 2022. It's now a reality.

---

𝙃𝙚𝙧𝙚'𝙨 𝙝𝙤𝙬 𝙩𝙝𝙞𝙨 𝙩𝙧𝙖𝙣𝙨𝙡𝙖𝙩𝙚𝙨 𝙞𝙣𝙩𝙤 𝟴 𝙠𝙚𝙮 𝙨𝙩𝙚𝙥𝙨:

1. Define purpose & scope

Over-scoping is the killer of agents. You want a single clear use case with success criteria, and constraints.

2. Build system prompt

This is often the most important step. You want a really good prompt, or the whole thing falls over.

3. Choose LLM

Don't always go for the best. Balance speed, multi-modality, and what each LLM is best for.

4. Integrate tools

This is what makes your agent an agent, MCP, APIs, other agents as tools. Create your architecture.

5. Add memory

Don't add memory unless needed. But for support agents that need conversation history, or research agents that build on past findings? Game-changer.

6. Orchestrate things

You want a control flow over routes, workflows, message queues and your agent-to-agent (A2A) coordination.

7. Build a user interface

This is the flashy part we want to jump to. But only it do it now. Figure out how users will interact.

8. Add testing & evals

This is the most important step to create a continuous improvement loop. Systematically analyze errors.

Here's guides to help with each step:

1. Purpose ➝‬ https://lnkd.in/eekC6bwT
2. System prompt ➝‬ https://lnkd.in/d_qYCBT7
3. Choose LLM ➝‬ https://lnkd.in/eqRgsMZx
4. Integrate tools ➝‬ https://lnkd.in/ezNsTfbD
5. Add memory ➝‬ https://lnkd.in/eeey5Cxr
6. Orchestrate things ➝‬ https://lnkd.in/eKhmKYNp
7. Build a UI ➝‬ https://lnkd.in/epRC64ku
8. Evals ➝‬ https://lnkd.in/eGbzWMxf

---

𝙏𝙚𝙘𝙝 𝙨𝙩𝙖𝙘𝙠:

Now, what tools should you use?

For general assistants: ChatGPT, Claude, Perplexity
For coding projects: Cursor, Windsurf, Claude Code
For business automation: Lindy, Relay app, n8n
For complex workflows: LangGraph, CrewAI, LlamaIndex

The best stack is the one that gets you shipping fast.

---

Agents are one of the most powerful skills for PMs and engineers to learn right now.

I have been studying them deeply.
These are my core learnings.

📌 Want my 8-layer agent architecture? Comment 'agent architecture' + DM me.

What agents have you been building?

I just saw the future of product building.

Brainstorm to 3 prototypes with AI:

The workflow used to be:
1. Brainstorm
2. Write a prompt/PRD
3. Build in a prototyping tool

(And still is at most companies)

What Miro unveiled at Canvas 25 is:

1. Brainstorm
2. Miro AI uses the canvas as the prompt
3. Miro Prototypes generates 3 diverse prototypes

Check it out: http://miro.pxf.io/9L0EqY

The best way to reduce product risk is prototypes. This gets you there much faster. It's AI at its best.

You can take this one step further with formats.

You can create:
• Docs
• Slides
• Tables
• Diagrams

Just using the canvas as context engineering.

Here's a full breakdown: http://miro.pxf.io/xLYZo3

The future of AI isn't heavier prompts. It's pulling context from what we're already doing.

PMs are busy enough. This takes work off your plate.

#ProductManagement #MiroPartner #Canvas25

Want to become an AI PM who earns $300K+?

I spent 100s of hours creating an email course to help you:

→ Understand the fundamentals of AI PM
→ Build AI agents that enhance your workflows
→ Develop the critical skills you need to ace this role

(From context engineering to AI evals)

Want access?

📌 📌 Update: Here is where you can find the email course for free: aipmcourse.com

No need to leave a comment or send me a connection request now.

ChatGPT & Lovable are just the tip of the iceberg.

PMs should use 10 buckets of AI tools:

𝗗𝗜𝗦𝗖𝗢𝗩𝗘𝗥𝗬

𝟭. 𝗣𝗿𝗼𝘁𝗼𝘁𝘆𝗽𝗶𝗻𝗴
Yes, this is an amazing PM use case.

Tools like Lovable, Bolt, v0, Magic Patterns, and Base44 are great for improving your discovery work.

🔗 Guide: https://lnkd.in/eJujDhBV

𝟮. 𝗖𝘂𝘀𝘁𝗼𝗺𝗲𝗿 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲
As important as prototyping, these aggregate insight.

Tools like Dovetail, Enterpret, Unwrap, Monterey, and Sprig make interpreting 10,000+ data points easier.

🔗 Guide: https://lnkd.in/evCwj4rU

𝗗𝗘𝗟𝗜𝗩𝗘𝗥𝗬

𝟯. 𝗩𝗶𝗯𝗲 𝗖𝗼𝗱𝗶𝗻𝗴
Some PMS you go beyond prototype to actually coding.

Tools like Cursor, Claude Code, OpenAI Codex, Replit, and Warp let you ship small + certain features fast.

🔗 Guide: https://lnkd.in/e66DrW-h

𝟰. 𝗩𝗶𝗯𝗲 𝗘𝘅𝗽𝗲𝗿𝗶𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻
These days, you can prompt entire experiments.

Tools like Optimizely, Amplitude, Kameleoon, Pendo, and LaunchDarkly open up prompt-based experimentation.

🔗 Guide: https://lnkd.in/e86mpjGR

𝗣𝗥𝗢𝗗𝗨𝗖𝗧𝗜𝗩𝗜𝗧𝗬

𝟱. 𝗗𝗶𝗰𝘁𝗮𝘁𝗶𝗼𝗻
Voice-to-text everywhere.

Tools like Wispr, SuperWhisper, Apple, TalkType, and Speechify are save hours.

🔗 Demo: https://lnkd.in/epRC64ku

𝟲. 𝗚𝗲𝗻𝗲𝗿𝗮𝗹 𝗟𝗟𝗠𝘀
You need to use the right LLM for the right thing.

Claude is the best writer, NotebookLM sole-context, Veo 3+Nano Bana for Image/Video, and GPT-5 are the foundation.

🔗 Guide: https://lnkd.in/eZ6UBbAQ

𝟳. 𝗠𝗲𝗲𝘁𝗶𝗻𝗴𝘀
Meeting next steps have never been easier.

Tools like Granola, Fathom, Otter.ai, Tldv, and Fireflies do 90% of the work.

🔗 Guide: https://lnkd.in/eMiacJW3

𝗔𝗚𝗘𝗡𝗧𝗦

𝟴. 𝗔𝗜 𝗖𝗼𝗱𝗶𝗻𝗴
PMs now use automonous development agents that ship code for bugs & small + certain features.

Tools like Linear, CodeGen, Devin, Sweep, and Codium lead the way.

🔗 Demo: https://lnkd.in/eKuqyc-g

𝟵. 𝗙𝘂𝗹𝗹-𝗙𝗲𝗮𝘁𝘂𝗿𝗲𝗱 𝗔𝗴𝗲𝗻𝘁 𝗣𝗹𝗮𝘁𝗳𝗼𝗿𝗺𝘀
Advanced workflow orchestration for complex stuff.

Tools like n8n, make.com, Activepieces, Workato, and Tray.io are handling the hard automations.

🔗 Demo: https://lnkd.in/e9QrMEDw

𝟭𝟬. 𝗦𝗶𝗺𝗽𝗹𝗲 𝗔𝗴𝗲𝗻𝘁 𝗣𝗹𝗮𝘁𝗳𝗼𝗿𝗺𝘀
No-code automation builders for quick wins.

Tools like Zapier, Lindy, Relay, Bardeen, and Parabola are connecting apps easily.

🔗 Zapier CTO: https://lnkd.in/dVp93fRE
🔗 Lindy CEO: https://lnkd.in/e7_YSFUw
🔗 Relay CEO: https://lnkd.in/e6qr34J3

𝗟𝗲𝗮𝗱𝗲𝗿𝘀 + 𝗜𝗧 𝗱𝗲𝗽𝗮𝗿𝘁𝗺𝗲𝗻𝘁𝘀: 𝗴𝗲𝘁 𝗹𝗶𝗰𝗲𝗻𝘀𝗲 𝗳𝗼𝗿 𝘆𝗼𝘂𝗿 𝗣𝗠𝘀!

PMs need each of these buckets, not just Copilot.

𝗧𝗵𝗲𝘀𝗲 𝗮𝗿𝗲 𝗺𝘆 𝘁𝗼𝗽 𝟱𝟬 𝗔𝗜 𝘁𝗼𝗼𝗹𝘀 𝗳𝗼𝗿 𝗣𝗠𝘀.

What would you add or remove?

Director at Booking: I HATE the title AI PM!

Here's what's going on:

The title AI PM is a very umbrella term.

It obscures several different types of roles within AI PM.

There are at least four major types that are different:

1. AI Research PMs
2. AI Products PMs
3. AI Platform PMs
4. AI Powered PMs

1. AI Research PMs

These are the PMs building LLMs, working with frontier AI research teams at OpenAI, Google, Anthropic...

They need to be masters of:

LLM building: https://lnkd.in/eeqTmeDU
Latest in AI research: https://lnkd.in/erfrE-KN
Integrity: https://lnkd.in/eMPvzcts

2. AI Products PMs

These are the PMs building AI features on top of LLMs, at almost every company in the world, but also in the frontier labs and AI companies like Cursor as well.

They need to be experts in:

Prompt engineering: https://lnkd.in/d_qYCBT7
Context engineering: https://lnkd.in/ebfnDUmi
Evals: https://lnkd.in/eGbzWMxf

3. AI Platform PMs

These are the PMs building the infrastructure, tooling, and systems that enable AI products to scale from prototype to production.

MCP: https://lnkd.in/e9Eh-KPq
AI Observability: https://lnkd.in/e3eQBdMp
Model deployment: https://lnkd.in/efcv-DXZ

4. AI Powered PMs

These are literally every other PM that doesn't need to build AI features but needs to use AI tools for productivity.

They need to master:

AI Prototyping: https://lnkd.in/eJujDhBV
AI Agents: https://lnkd.in/eeey5Cxr
+ More: https://lnkd.in/eCFUPQtM

Marily Nika said there are 11 types: https://lnkd.in/e3akMRm8

In his post, Pranav talks about 6 types.

There's a million ways to break it down.

Suffice to say, AI PM is an umbrella term.

📌 Want my deeper breakdown into types of AI PM? Comment 'AI PM types' + DM me (free inMail).

What type of AI PM do you want to become?

16 awesome resources that will get you hired as a Product Manager 👇

Whether you're interviewing at Big Tech or a startup, this list covers what you actually need to stand out - from writing a resume to tackling product strategy questions and behavioral interviews.

1. PM interview preparation tips from Amazon:
🔗 https://lnkd.in/e_BJxUzN

2. Guide to mastering product sense interviews:
🔗 https://lnkd.in/eAGzY74j

3. Database of 150+ Big Tech PM interview questions:
🔗 https://lnkd.in/eKshXQEA

4. Guide to PM metrics interview questions:
🔗 https://lnkd.in/eeJvYHNb

5. How to Ace Tell Me About Yourself:
🔗 https://lnkd.in/evX2VNXq

6. PM case studies from top companies to practice:
🔗 https://lnkd.in/eB2k6xRE

7. PM interview guide for breaking into cos like Google:
🔗 https://lnkd.in/eDaGV38Y

8. Flip weaknesses method to nail behavioral interview:
🔗 https://lnkd.in/egFdtXhQ

9. How to nail the job search:
🔗 https://lnkd.in/ezYPrW2t

10. Product design interview breakdown:
🔗 https://lnkd.in/gWjKm5x8

11. Resume building guide with FAANG templates:
🔗 https://lnkd.in/e8sRUfDY

12. Meta's guide to PM interviews:
🔗 https://lnkd.in/eW6NhRpS

13. Small market recruiting strategy:
🔗 https://lnkd.in/evQeJZGB

14. PM Interview Cheat Sheets:
🔗 https://lnkd.in/eRiDFXZi

15. Guide to Metrics Interviews:
🔗 https://lnkd.in/exQUkvTh

16. Guide to Break into AI PM Roles:
🔗 https://lnkd.in/e_M2EJ6v

Bonus: Want to get my coaching?
➤ Join my cohort: https://www.landpmjob.com/

Google, Meta, OpenAI and the rest are hiring like crazy.

📌 Want my free 7 day AI PM course? Comment 'EEC' + DM me.

You should vibe code a PM portoflio.

Here's how to make a great one:

Only 17% of PMs have a portfolio. So you should differentiate with one:

1. Headline

This is the most important part! You want to describe your unicorn candidate-market fit. This is your opportunity to make a bumpy career look like a straight line to a sector of the market.

2. Navigation

The beauty of a website is you can add layers. Add in followup pages to deep dive into your experience. Link them. Add in blog & podcast appearances too.

3. Picture

This is more important than you might think! People shouldn't care, but they do. Have a professional, recent photo that looks like you.

4. Career Highlights

You want to go beyond LinkedIn and the resume. Give a clear summary of what you did with quotes. Make it pop.

5. Work Products

Your resume and LinkedIN, people assume, is full of half-truths. Work products speak louder than words. They show the quality of what you can produce.

6. Help Them Reach Out

Make it easy for people to download your resume, email you, or connect on LinkedIn. Drive to an outcome!

7. End with Fit

You want to re-summarize your unicorn candidate market fit so people know what you are + aren't a fit for.

8. Bonus

Add in things to make it pop:

• Cover weaknesses with work products
• Add in a video sales letter about yourself
• Show real work products that you can open source
• Put in personal touches like your passions

Want my prompt to build your portfolio?

Full guide: https://lnkd.in/eT3RvQiM

Put that together with these pieces:

• Theory on portfolios: https://lnkd.in/eT3RvQiM
• How to make your resume: https://lnkd.in/ej-2bZVb
• Optimize your LinkedIn: https://lnkd.in/egeXb7Sc
• AI PM job search: https://lnkd.in/efbppEDj

And you'll land your dream job in no time.

📌 For my recommendation of which tool to use, and how to use it, comment 'portfolio tool' + DM me.

Want my coaching on this? Join my cohort starting in 10 days. Only 17/30 seats left: www.landpmjob.com

Have you vibe coded your portfolio? Link it below 👇

9 out of 10 AI tools are hype.

These are the one's AI PMs should care about:

I sat down with Anshumani Ruddra to tier all the tools.

Anshumani was the person to do this with. He's a GPM at Google responsible for all APAC payments and on the BLEEDING edge of AI tool use.

He's the guy running 6 parallel Claude Code windows.

🎬 Watch now: https://lnkd.in/e9uzHb57

Spotify: https://lnkd.in/eyt7agKj
Apple: https://lnkd.in/eVZf64gB

🏆 Thanks to our sponsors:

1. Miro: The AI innovation workspace - https://lnkd.in/esuXdmv3
2. Vanta: Leading AI security & compliance platform - http://vanta.com/aakash
3. Testkube: Top test orchestration platform - http://testkube.io/
4. Kameleoon: Leading AI experimentation platform - http://www.kameleoon.com/
5. Dovetail: The AI customer intelligence platform - https://dovetail.com/

Here's all the tools we cover:

-AI Agents: n8n, make dot com, Lindy, Airtable, Zapier, Relay app
-AI prototyping: Lovable, Bolt, Magic Patterns, v0, Base44
-Vibe coding: Replit, Windusrf, Claude Code, Cursor, Codex, Github Copilot
-LLM: ChatGPT, Perplexity, Claude Sonnet, Grok, Gemini, Microsoft Copilot, Manus
-Data Analysis & Insights: Amplitude, Kameleoon, Optimizely, Statsig
-Discovery & research: Unwrap AI, Enterpret, Dovetail
-Browser: Perplexity Comet, Dia
-Roadmapping: Jira Product Discovery, Pendo, Productboard
-Communication & Documents: Notion, Jasper, Copy ai, Grammarly, Gamma
-Meetings: Granola, Fireflies ai, Otter ai, Fathom, tl;dv
-Dictation: wispr flow, superwhisper
-AI Video: Descript, Loom, Synthesia, Kling
-AI Design: Figma AI, Uizard
-Project Management: Linear

You'll have to watch the video for all the rankings!

But our top S-Tier top tools were:

1. Claude Code: not just a coding tool, but the best
2. superwhisper: the best dictation tool
3. Lindy: top AI agent builder
4. Replit: top prototyper

What are your top tools?

He runs a $10M rev/year company with AI (1 employee)

It's a preview into the future of PM:

While most PMs live with rigid 2-week sprints, write long PRDs, and spend all their time coordinating people...

Sahil Lavingia ships ideas to production in 10 minutes.

While working on 5 things at a time.

Today's episode breaks down how you can:

🎬 Watch Now: https://lnkd.in/eE-CjMU3

Spotify: https://lnkd.in/eyt7agKj
Apple: https://lnkd.in/eAEVwr3u

🏆 Thanks to our sponsors:

1. Vanta: Get $1,000 off AI security & compliance - http://vanta.com/aakash
2. Testkube: Leading test orchestration platform - http://testkube.io/
3. Kameleoon: Leading AI experimentation platform - http://www.kameleoon.com/
4. The AI PM Certificate: Get $550 off with ‘AAKASH550C7’- https://lnkd.in/eZ2uUySG

This isn't hype.

As he said, "these models are as smart as they need to be to replace most software engineering."

Today's episode is a tutorial to using Devin to ship to production.

He covers 3 different scenarios:

1. Small, well spec'd item → straight to Devin
2. More complex item → discussed with team on Github
3. Big, uncertain feature → prototyped in v0

This is really the future of product building. Junior eng will be replaced, and the PRD will not exist for simple + certain features.

Of course, it will take a long time in big companies where PRDs exist for coordination, but...

"The PRD is kind of dying"

- Sahil

What do you think about the future of the PRD?

I wrote about it here:

PRD History: https://lnkd.in/ef5Yde54
Future of AI PRDs: https://lnkd.in/eMu59p_z
AI Prototyping + PRDs: https://lnkd.in/eJujDhBV

📌 Comment 'Sahil workflow' + DM me for a step-by-step of his futuristic workflow.

PS. You don't want to miss why he doesn't own ANY public stocks.. despite making $2M/year!

It was such a treat to record with Sahil. I've been following him on X for years. I saw him in person in 2021 & was too shy to talk to him. This. was. awesome.

Vibe code → production?

It's finally possible now:

Maor Shlomo did it.

Sold for $80M cash with an app he mainly vibe coded.

𝗧𝗵𝗶𝘀 𝗶𝘀 𝗮 𝗰𝗿𝗶𝘁𝗶𝗰𝗮𝗹 𝘀𝗸𝗶𝗹𝗹 𝗳𝗼𝗿 𝗳𝗼𝘂𝗻𝗱𝗲𝗿𝘀 𝗮𝗻𝗱 𝗣𝗠𝘀

"Anyone who announces that vibe coding can't lead to production ready software is just telling on themselves."

The latest Claude 4.5-Sonnet model in claude code or Cursor can generate production code.

𝘏𝘦𝘳𝘦 𝘢𝘳𝘦 𝘵𝘩𝘦 𝘴𝘵𝘦𝘱𝘴 𝘵𝘰 𝘥𝘰𝘪𝘯𝘨 𝘪𝘵.

Full Guide: https://lnkd.in/eNB4mQY7

(Repost + reshare the summary for others ♻️)

𝗦𝘁𝗲𝗽 𝟭: 𝗙𝗶𝗻𝗱 𝗬𝗼𝘂𝗿 𝗜𝗱𝗲𝗮

Write a great PRD. This is the first step:

• What problem you are solving
• How the metrics will work
• The user flows

🔗 Guide: https://lnkd.in/eMu59p_z

𝗦𝘁𝗲𝗽 𝟮: 𝗦𝗮𝗮𝗦 𝗧𝗲𝗺𝗽𝗹𝗮𝘁𝗲 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻

You don't want to start with bad vibe code.

If you start with a strong SaaS template, you'll have a sturdy foundation to go to production.

𝗦𝘁𝗲𝗽 𝟯: 𝗗𝗮𝘁𝗮 𝗠𝗼𝗱𝗲𝗹𝗶𝗻𝗴

People want to outsource and vibe this. But it's the most critical step.

With a bad data model, your app collapses when people actually use it. Sketch out the data schema.

𝗦𝘁𝗲𝗽 𝟰: 𝗖𝗼𝗺𝗽𝗼𝗻𝗲𝗻𝘁𝘀 & 𝗩𝗶𝘀𝘂𝗮𝗹 𝗗𝗲𝘀𝗶𝗴𝗻

Using Tailwind, you can define the design system really easily.

This gives the AI all the building blocks to build your app the way you want it to like and not the same as all other vibe coded apps.

🔗 Detail: https://lnkd.in/ejDETBtk

𝗦𝘁𝗲𝗽 𝟱: 𝗣𝗿𝗼𝘁𝗼𝘁𝘆𝗽𝗶𝗻𝗴 & 𝗩𝗮𝗹𝗶𝗱𝗮𝘁𝗶𝗼𝗻

Start with a prototyping tool to get a good testing version.

Use this in discovery efforts to nail down your core MVP UX.

🔗 Guide: https://lnkd.in/eJujDhBV

𝗦𝘁𝗲𝗽 𝟲: 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗜𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻

Then you're ready to go construction mode. Go from prototyping tool to Cursor + Claude Code.

You're going to want to build out the full app.

𝗦𝘁𝗲𝗽 𝟳: 𝗖𝗜/𝗖𝗗 𝗣𝗶𝗽𝗲𝗹𝗶𝗻𝗲

Set up automated deployment so you're not manually pushing code like it's 2015.

GitHub Actions or Vercel make this trivial. Every commit goes through tests, builds, and deploys automatically.

𝗦𝘁𝗲𝗽 𝟴: 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗘𝘀𝘀𝗲𝗻𝘁𝗶𝗮𝗹𝘀

Authentication, rate limiting, API security. The basics that keep you from getting hacked on day one.

Use established libraries for auth (Clerk, Auth.js).This is where you follow best practices, not reinvent.

𝗦𝘁𝗲𝗽 𝟵: 𝗗𝗶𝘀𝘁𝗿𝗶𝗯𝘂𝘁𝗶𝗼𝗻

Ship it. Get it in front of users immediately.

Launch on Product Hunt. Post on X. Share in relevant communities. The best validation is real people using it.

You can't vibe code product-market fit: you find it by shipping and iterating.

📌 Want my starter prompt? Comment 'Start' + DM me.
➕ Follow Aakash Gupta for daily AI PM insights.

The PM role has changed.

Now, every PM needs a prompt library:

Leaders are even checking it before hiring.

So you want to cover all the major use cases:

GTM
Career
Analytics
DIscovery
Operations
AI Features
Productivity
PM Artifacts
Strategy & Planning

And you want to use the latest prompting techniques.

Since that's a lot, I created one for you.

Check it out: https://lnkd.in/eHU88esH

I spent 2 years building this library.

Then I hired a researcher who helped me:

a. Harden all these prompts
b. Test them against other libraries
c. Make sure they don't hallucinate
d. Make it easy to input context in <60s

Grabbing it is like hiring 82 little junior PM interns.

You transform LLMs from chatbots into APMs.

📌 Want the mind map of prompts you should have in your library? Comment 'prompt mind map' + DM me.

Drop your favorite prompts below 👇