The Product Growth Guy 🚀 | Helping you land your next job + succeed in your career | Newsletter Writer and Podcast Host

TL;DR: I write a newsletter for product managers, product leaders, and aspiring product managers. I focus on actionable and deep tips to improve your career.

The Longer Version:

After a long career in PM - rising to VP of Product at a Unicorn - I have made the shift to writing about it.

In particular, I write about product growth. That is, the art of growing via your product. 

This means I also talk a lot about product-led growth (PLG), the SaaS B2B motion.

My newsletter doesn't just focus on growth though. 

I also break down technology, product management, product leadership, and how to get a PM job. 

It's all the things that make for a great use of your learning & development budget. 

Or even personal spend. 

The math is pretty great. 
↳ $150 for an annual subscription
↳ $211K average PM salary in the US
↳ 0-10% average range of raises offered

If the newsletter helps you get a 0.07% better raise, it pays itself. 
That's 1/142th of your raise range.

It's a pretty smart investment for every level of PM.

http://www.news.aakashg.com

Aakash Gupta

Claude Code is the most powerful AI today.

(Especially with yesterday's Opus-4.6 release)

Here's why it reached $1B ARR in 6 months and how to master it:

1. It changes everything

Full system access, agent swarms, hours-long tasks. Chat Claude is browser-only. Cowork is a limited skin.

Guide to Claude: https://lnkd.in/eZ6UBbAQ 
Guide to Cowork: https://lnkd.in/g6MhB3YJ

2. What Claude Code can do

Process 10,000 files in minutes. Run bash without learning syntax. One interface for 200+ tools. Delegate like you have 5 junior employees.

Basic Guide: https://lnkd.in/eUyPEAma 
Advanced Guide: https://lnkd.in/gvA6rZtX

3. The Claude Code workflow

Analyze → Plan → Create → Scale

Meeting transcript → PRD → Jira tickets → Slack summary → dashboard. One tool.

AI PRD Guide: https://lnkd.in/eMu59p_z 
AI Roadmap Guide: https://lnkd.in/eEuNcj-m

4. Getting started

Use Cursor. Open terminal, type "claude", press Enter.

macOS 13+, Ubuntu 20+, or Windows 10+. Pro ($20/mo), Max ($100-200/mo), or API.

Pro Guide: https://lnkd.in/eadQ5fN6 
Official Setup: https://lnkd.in/gPthuahX

5. Connect everything (MCP)

USB-C for AI. One protocol, hundreds of connections. GitHub, Sentry, Linear, Notion, Slack, Jira, PostgreSQL, Gmail, Typefully, Buffer, Discord.

PM Operating System: https://lnkd.in/gsH9RSfJ
Anthropic MCP Servers: https://lnkd.in/g-sx7vXP 
MCP Collection: https://mcp.so

6. Essential commands

/help → all commands 
/clear → reset context 
/compact → save tokens 
/model → switch models 
/mcp → check connections

Use @ to reference files (@filename, @folder/).

Prompt Engineering: https://lnkd.in/d_qYCBT7 
Best Practices: https://lnkd.in/gNSNWmKB

7. Skills & CLAUDE.md

Skills are reusable automations Claude auto-loads. Built-in for docx, xlsx, pptx, pdf.

CLAUDE.md is project memory. Commands, style guides, context that persists.

Skills Tutorial: https://lnkd.in/eRidpGSR 
Steal 6 of My Skills: https://lnkd.in/eZyJzMnH 
Context Engineering: https://lnkd.in/eUUPMmJK

8. Prompting techniques

Pro pattern: Checkpoint + Iterate.

Plan first. Review before execution. Create checkpoints. Rewind (Esc twice) if it breaks. Iterate with specific feedback.

Prompt Engineering: https://lnkd.in/d_qYCBT7 
ChatGPT for PMs: https://lnkd.in/eyfUi8ZN

9. Resources

All links above
Podcast: https://lnkd.in/eyt7agKj
MCP Reference: https://lnkd.in/gPUWGa7S

🔖 Save this. Repost so your network sees it too.
➕ Follow Aakash Gupta for daily AI tips.

Claude code is the future. 

This is your guide.

Helping you succeed in your career + land your next job

Anthropic PMs don't write PRDs.

They use Claude Code to build the first version of the feature themselves. The whole company dogfoods it for weeks. Then they ship an experiment.

Anthropic went from $1B run-rate revenue at the start of 2025 to over $9B by year end. Claude Code alone generates over $500M in run-rate revenue with usage growing 10x in three months. The company signed a term sheet at $350B, nearly doubling from $183B just months earlier.

This is what happens when you eliminate handoff drift. In a traditional org, a PM writes a spec. Design interprets the spec. Engineering interprets the design. Each handoff introduces drift. Each handoff adds a week. The 2-4 week cycle of PRD → design → ticket → sprint → build → QA compresses to days when the PM can prototype in the codebase.

The prerequisite is making your codebase AI-friendly. Tailwind instead of scattered CSS files. Microservices architecture. Reduced codebase size. That infrastructure investment is what lets PMs vibe code directly against production components.

The math is simple. If your experimentation cycle is 8 weeks per test and your competitor's cycle is 3 days, they're running 13x more experiments per quarter. Compound that over a year and the learning gap becomes insurmountable.

For more on this way of working:

1. My Berkeley keynote: https://lnkd.in/g6tNMu2j
2. Claude code basics: https://lnkd.in/gij3jSvH
3. Claude code advanced: https://lnkd.in/gr9sPtJ7
4. Claude code OS: https://lnkd.in/gsH9RSfJ
5. AI PRDs: https://lnkd.in/eMu59p_z
6. AI prototyping: https://lnkd.in/eJujDhBV

There’s places where this doesn’t work. But they’re less than most people realize. And the benefits are tough to argue.

The PM skill that matters in 2026 is taste at speed. Boris Cherny just showed everyone what that looks like.

His Claude Code team at Anthropic doesn’t write PRDs. They build hundreds of working prototypes before shipping a single feature. Boris personally ships 20-30 PRs a day running 5 parallel Claude instances. They built Cowork, a full product for non-engineers, in about 10 days.

Everyone is debating whether PRDs should die. They shouldn’t in big companies and established products, where understanding the hypothesis and success metrics is critical. But the real question is: what happens to the PM who can’t evaluate 15 prototypes and pick the 3 worth shipping?

Because here’s what changes when building costs near zero: the bottleneck moves from “can we build it” to “should we ship it.” PRDs existed because building was expensive and you needed sign-off before committing resources. When a prototype takes 45 minutes instead of 6 weeks, nobody needs a document to authorize exploration. They need someone who can look at working software and say “this one, not that one” in real time. If anything, the PRD now comes after the prototype.

On the Claude Code team, PMs code. Data scientists code. User researchers code. In fact, now everyone has the same title (“Member of Technical Staff”), and it’s by design. Boris said productivity per engineer grew 70% even as Anthropic tripled in headcount. The coordination cost of translating specs into code disappears when everyone can build. And that changes what a PM is actually good for.

Boris said it himself: “There’s just no way we could have shipped this if we started with static mocks and Figma or if we started with a PRD.” The old process would have spent more calendar time documenting Cowork than his team spent building it.

This is the Claude Code team today. It will be most fast-moving teams within 18 months. The PMs who thrive will be the ones reviewing prototypes at 9am, killing 80% of them by noon, and shipping the survivors by end of week. Pattern matching across user research, technical feasibility, and business model simultaneously while staring at working software.

So here's how to get there:

1. Understand Modern PRDs: https://lnkd.in/eMu59p_z
2. Master Claude Cowork: https://lnkd.in/gwhgmSfw
3. Learn Claude Code: https://lnkd.in/eUyPEAma
4. Build your OS: https://lnkd.in/gsH9RSfJ
5. Watch Boris interview: https://lnkd.in/g8fXRPRR

The PMs who struggle will be the ones still writing 15-page specs for features that could be prototyped, tested, and validated before the doc hits its first review cycle. Taste at speed is the new moat.

No alternative text description for this image

The math on this project should mass-humble every AI lab on the planet.

1 cubic millimeter. One-millionth of a human brain. Harvard and Google spent 10 years mapping it. The imaging alone took 326 days. They sliced the tissue into 5,000 wafers each 30 nanometers thick, ran them through a $6 million electron microscope, then needed Google’s ML models to stitch the 3D reconstruction because no human team could process the output.

The result: 57,000 cells, 150 million synapses, 230 millimeters of blood vessels, compressed into 1.4 petabytes of raw data. For context, 1.4 petabytes is roughly 1.4 million gigabytes. From a speck smaller than a grain of rice.

Now scale that. The full human brain is one million times larger. Mapping the whole thing at this resolution would produce approximately 1.4 zettabytes of data. That’s roughly equal to all the data generated on Earth in a single year. The storage alone would cost an estimated $50 billion and require a 140-acre data center, which would make it the largest on the planet.

And they found things textbooks don’t contain. One neuron had over 5,000 connection points. Some axons had coiled themselves into tight whorls for completely unknown reasons. Pairs of cell clusters grew in mirror images of each other. Jeff Lichtman, the Harvard lead, said there’s “a chasm between what we already know and what we need to know.”

This is why the next step isn’t a human brain. It’s a mouse hippocampus, 10 cubic millimeters, over the next five years. Because even a mouse brain is 1,000x larger than what they just mapped, and the full mouse connectome is the proof of concept before anyone attempts the human one.

We’re building AI systems that loosely mimic neural networks while still unable to fully read the wiring diagram of a single cubic millimeter of the thing we’re trying to imitate. The original is 1.4 petabytes per millionth of its volume. Every AI model on Earth fits in a fraction of that.

The brain runs on 20 watts and fits in your skull. The data center required to merely describe one-millionth of it would span 140 acres.

If you want to stay up on AI, do not miss my AI newsletter: www.aibyaakash.com

I don't think most PMs realize how fast the "AI-native PM" bar is rising.

Six months ago, knowing how to write a good prompt was enough. Now the PMs getting hired at top AI companies have full tooling stacks: custom GPTs for PRD drafts, Claude Projects loaded with their company's design principles, Gemini Gems configured for competitive analysis, Claude Code OS with context.

Lisa Huang created Gems at Google. She showed her complete setup on the podcast and the pattern is clear. The best PMs aren't prompting AI. They're configuring AI. Once.

The setup cost is 2 hours. The ongoing return is 5+ hours per week. The compounding return is that your AI gets better at your job the more you use it.

Every PM interview I've seen in 2026 asks some version of "how do you use AI in your workflow." The answer they want isn't "I use ChatGPT." They want to hear about your system.

Here's how to ace those questions:

1. Watch Lisa's ep on Gems: https://lnkd.in/gbTJTt_3
2. Learn Claude Projects: https://lnkd.in/eZ6UBbAQ
3. Grab my OS: https://lnkd.in/gsH9RSfJ
4. Grab my prompt library: https://lnkd.in/eHU88esH
5. Practice your interviews: https://lnkd.in/eZZ_vvQc

Build the system.

The creator of Claude Code, which hit $1B ARR faster than almost any dev tool in history, just told you the endgame: the title “software engineer” gets replaced by “builder” or “product manager.”

This is Anthropic’s head of Claude Code saying his own product eliminates the skill that defined his entire career. Cherny rose to Principal Engineer (IC8) at Meta over seven years. He wrote and reviewed code for a living. Now he runs 10-15 concurrent Claude sessions and treats AI like compute you schedule, not a tool you prompt.

In a podcast with YC, Cherny posted that engineers are “more important than ever.” When someone pointed out the contradiction with Dario’s Davos prediction that models will do all of SWE end-to-end in 6-12 months, Cherny replied: “Dario is talking about what’s next.”

This tells you everything about Anthropic’s internal positioning. They’ve gone from $1B to $7B revenue run rate in nine months. They have 50+ engineering roles open right now. And the guy building the tool that automates coding is simultaneously hiring more engineers than ever.

The contradiction dissolves when you realize what those engineers actually do. Cherny uses a “verify-app” agent to run end-to-end tests, a “code-simplifier” agent for architecture cleanup, and a PostToolUse hook that auto-formats the last 10% Claude misses. He’s building systems that make AI reliable, not writing application code.

Anthropic is betting that the value shifts from people who can write code to people who can define what to build, sequence the work across parallel agents, and catch the 10% of failures that compound into production disasters. That’s the “builder” Cherny is describing.

5 resource to move to this way of working:

1. New tool stack: https://lnkd.in/gtJWybA5
2. Claude code for PMs: https://lnkd.in/eUyPEAma
3. Advanced claude code: https://lnkd.in/eadQ5fN6
4. Claude code PM OS: https://lnkd.in/gsH9RSfJ
5. Boris' interview: https://lnkd.in/ge2JGgCd

Every PM reading this should understand: the person who built the most successful AI coding tool in history just said your job title is the future of software. The question is whether you’re building the judgment and technical fluency to claim it.

If I were to learn PM again, I would start here. 

15 steps to a PM job paying $200K+:

1. Understand the PM role

Start with how to break in: https://lnkd.in/erSRWwUF
Watch Marty Cagan: https://lnkd.in/eb5hbA28

2. Learn PM fundamentals

Master PRDs: https://lnkd.in/ef5Yde54
Watch Dan Olsen: https://lnkd.in/gWgUjVpB

3. Master product strategy

It's the key PM skill: https://lnkd.in/e8EUQHAW
Hear from Roger Martin:  https://lnkd.in/ecVF-y6i

4. Learn product discovery

Master continuous discovery: https://lnkd.in/eYsTAnrq
Hear Teresa Torres: https://lnkd.in/gdVqf2bV

5. Understand growth

Master growth models: https://lnkd.in/dEfhHTMP
Hear from Brian Balfour: https://lnkd.in/eef_tipn

6. Learn analytics & experimentation

Master metrics: https://lnkd.in/ea8sWSsS
Watch Amplitude's CEO: https://lnkd.in/esuAeUUK

7. Study company growth case studies

Start with Cursor: https://lnkd.in/eq8xvCAK
Then read Linear: https://lnkd.in/e_R9ruDG

8. Build your PM Portfolio

Start with basics: https://lnkd.in/eQd32YUA
Then vibe code: https://lnkd.in/eT3RvQiM

9. Optimize Resume & LinkedIn

Learn to customize: https://lnkd.in/eyMAjsNc
Make LinkedIn your billboard: https://lnkd.in/egeXb7Sc

10. Master PM interviews

Learn each interview: https://lnkd.in/e8WkHDbK
Hear from Lewis Lin: https://lnkd.in/e4MkWiyD

11. Execute your job search

Learn how to get interviews: https://lnkd.in/eqn5Eqq6
Hear from Phyl Terry: https://lnkd.in/eVpXWuRN

12. Target companies

Build a company list: https://lnkd.in/eBCcuHy7

13. Nail team matching

Learn the process: https://lnkd.in/eG_YqMEn

14. Negotiate

Master negotiation: https://lnkd.in/e8ZJmvMe
Hear from Satish Mummareddy: https://lnkd.in/epwZ3Dfw

15. Grow your career

Learn product leadership: https://lnkd.in/ewTn7W6z
Hear the PM to VP journey: https://lnkd.in/g5wp7snd

🔗 All links: https://lnkd.in/g59eb3Zy

Want me to DM them to you? Comment 'PM roadmap' + DM me and I'll reply (repost to cut the line).

➕ Follow Aakash Gupta for daily PM content.

Replit's gross margins went from 36% to negative 14% in two months.

Same product. Same pricing. Same team. The only thing that changed: they launched a more autonomous AI agent that consumed more LLM resources than their pricing covered.

Traditional SaaS has 70-80% gross margins because one more subscriber costs almost nothing. AI products pay for compute on every prompt. Your best users are your most expensive users. That single fact breaks every pricing model designed for the SaaS era.

I mapped pricing across the top 50 AI startups by valuation with Moe Ali. Six patterns emerged.

The scariest finding: in most AI products, the P90 user costs 10-40x more than the P50 user. Both pay the same subscription. You're subsidizing your heaviest users with revenue from your lightest ones. And that subsidy grows as power users discover more ways to use the product.

Cursor learned this the hard way. They switched from flat 500 requests/month to a credit pool system. A developer burned the entire monthly allocation in a single day. $7,225 invoice. The CEO published a public apology on July 4th. The plan description quietly changed from "Unlimited" to "Extended" twelve days after launch.

Anthropic took a different approach. Their $17/$100/$200 tiers map to genuinely different user personas. A casual user, a power user, and a developer replacing an IDE. Those are different products with different willingness to pay. Then weekly rate limits targeting less than 5% of subscribers to push the heaviest users toward the API, where per-token pricing covers actual compute.

The pattern across all 50 companies: pure flat pricing is dying. Nearly half use two or three models simultaneously.

Here's the full breakdown:

1. Complete AI pricing guide: https://lnkd.in/gdKaQSMk
2. Replit guide: https://lnkd.in/gmA_c_AG
3. AI product strategy: https://lnkd.in/egemMhMF
4. AI agents guide for PMs: https://lnkd.in/eeey5Cxr

If you can't estimate your cost distribution across P10 to P90, you're not ready to set a price.

Only 24% of PM candidates have a GitHub.
 
Every PM I've placed at OpenAI, Anthropic, and Meta AI in the last year had one.
 
They didn't have 78K stars. They had three things: one project that solved a real problem, documentation that showed product thinking, and consistent activity.
 
Shubham Saboo went from Dev Rel to Senior AI PM at Google because his GitHub marketed him. Google reached out to him. When companies come to you, the interview-to-offer rate jumps from 22% to 37%.
 
But most PM GitHubs I review are anti-signals. Copy-pasted AI code with no tradeoffs section. Joke projects that prove nothing. Green square farms with 50 empty commits.
 
I put together the web's first guide on PM GitHubs:
 
• The 6 elements that made Shubham's GitHub get him hired
• Step-by-step setup with Cursor and Claude Code
• A starter kit repo you can fork today
• 50 project ideas mapped to PM skills
• The 6 mistakes I see candidates make
 
Full guide for paid subscribers:

Open article: This Github Got a PM Hired at Google by news.aakashg.com

This is tool stack of the best AI PMs I've talked to:

(You just need one in each category)

𝗕𝗨𝗜𝗟𝗗𝗜𝗡𝗚

Vibe Coding

• Cursor, Claude Code, Windsurf, Replit, or Warp
• You want to be able to build your own operating system
• My guide: https://lnkd.in/gsH9RSfJ

Prototyping
 
• Lovable, Bolt, v0, Magic Patterns, or Base44
• A prototype is a key pair to the PRD these days
• My guide: https://lnkd.in/eJujDhBV

𝗣𝗥𝗢𝗗𝗨𝗖𝗧𝗜𝗩𝗜𝗧𝗬
 
Dictation
 
• Wispr Flow, superwhisper, Tactiq, Speechify
• You talk 2x faster than you type. Use that
• My guide: https://lnkd.in/g3bRR7e6
 
Meetings
 
• Granola, Fathom, Otter.ai, tl;dv, or Fireflies
• Never take manual meeting notes again
• My guide: https://lnkd.in/gWk_t49C
 
General LLMs
 
• Claude, ChatGPT, Gemini, Kimi, or Grok
• Find your favorites and master prompt engineering
• My guide: https://lnkd.in/d_qYCBT7
 
𝗔𝗨𝗧𝗢𝗠𝗔𝗧𝗜𝗢𝗡
 
Simple Agents
 
• Zapier, Lindy, Relay, Bardeen, or Parabola
• Start here if you've never automated a workflow
• My guide: https://lnkd.in/eyamsW7a
 
Full-Featured Agents
 
• n8n, make, Activepieces, Workato, or Tray
• Graduate here when simple feels limiting
• My guide: https://lnkd.in/eeey5Cxr
 
𝗗𝗜𝗦𝗖𝗢𝗩𝗘𝗥𝗬
 
User Research
 
• NotebookLM, Perplexity, Elicit, Consensus, or Grain
• Use AI at every step to make it faster
• My guide: https://lnkd.in/gcD8_wNS
 
Customer Intelligence
 
• Dovetail, Unwrap, Enterpret, Monterey, or Viable
• Go from unstructured data to usable inputs
• My guide: https://lnkd.in/evCwj4rU

Want a pro version of each? 
↳ Grab my bundle: https://lnkd.in/eYxxVhZr

What would you add to the list?

A developer on GitHub just built a full development methodology for AI coding agents and it has 91,000 stars.

It’s called Superpowers, and it completely changes how your AI agent writes code.

The backstory is what makes this wild.

Jesse Vincent created Request Tracker in 1994. It became the most widely used open-source ticket tracking system on Earth. He ran the Perl programming language for three years. He co-founded Keyboardio and shipped ergonomic keyboards to 78 countries. He co-founded VaccinateCA during COVID and helped millions find vaccines.

Every one of those projects solved the same problem: helping people organize complex work they can’t hold in their heads.

Five months ago he released Superpowers, a set of markdown skill files that plug into Claude Code, Codex, or Cursor. The entire value is in the constraints. Before the agent writes a single line, it stops and asks what you’re building. It writes a spec in chunks small enough to read. It breaks implementation into 2-5 minute tasks with exact file paths. It enforces test-driven development so strictly that it deletes code written before tests exist.

91,000 stars in five months. For a repo that is mostly markdown files telling your AI agent to slow down.

That growth tells you something uncomfortable. The bottleneck in AI-assisted development right now is not model capability. The models are smart enough. The problem is they have zero discipline. They guess at specs, skip tests, and produce code you spend the next hour babysitting.

A guy who spent 30 years building systems for how humans organize work just built the system for how AI agents organize work. The career arc makes perfect sense in retrospect.

Resources:

1. Superpowers repo: https://lnkd.in/gN6NXzrM
2. My Claude Code OS: https://lnkd.in/gsH9RSfJ
3. Claude Code tutorial: https://lnkd.in/eXNvwvWw

The teams that figure out AI agent discipline first will look like they have a two-year head start. The teams that let agents freewheel will spend that time cleaning up.

Follow Aakash Gupta for daily AI + PM insights.

The psychology of why people fail with Claude Code is the same psychology behind why people buy gym memberships in January and quit by February.

The tool works. The person didn't build the habit infrastructure around it.

Claude Code is the most powerful professional software I've used. Cowork is right behind it. But both have the same adoption problem. The first experience determines everything. And the first experience with a blank setup is always mediocre because Claude is guessing about every variable that matters. Your role. Your product. Your metrics. Your writing preferences. What "good" looks like at your company.

The fix is a one-time setup that takes about 60 seconds.

A memory prompt that gives Claude your professional context. A CLAUDE.md file with your role, north star metrics, PRD structure, and writing rules that loads automatically every session. Skills that encode your exact workflows so quality stays consistent across dozens of uses.

Then you add the compounding loop. Every time Claude makes a mistake, you add a rule to the CLAUDE.md. Claude proposes the edit, you approve it, and next session it's already loaded. Your setup gets smarter with every use. Most people's stays static.

Cowork extends this into automation territory. Scheduled competitive intel reports. Meeting transcripts turned into action items in 30 seconds. Browser-based research using your actual logged-in Chrome session. Spreadsheet creation from messy files.

And one habit that separates people who stick with it from people who don't: /clear between unrelated tasks. Context bleed is the number one failure mode. You write a PRD then ask for a status update in the same session. The update comes back written like a PRD.

Here’s everything you need for this:

1. Claude code quickstart: https://lnkd.in/gRsTZ6wH
2. Claude cowork guide: https://lnkd.in/gwhgmSfw
3. My PM OS: https://lnkd.in/gsH9RSfJ
4. 90 min video: https://lnkd.in/eUyPEAma

This is all the infrastructure you need to build the habit. And not quit the gym by February.

Your PMs should be building in the codebase.
 
Not full features. But prototypes on your actual design system. Polish on simple front-end changes. Sometimes the initial version of a feature. If your product team doesn't have access to these three tools, you're already behind.
 
1. Claude Code
 
This is the most powerful way to use an LLM. It connects your PM directly into the codebase. Not a separate prototype environment. Your actual elements, your actual design system. PMs building on top of production components, doing polish on simple front-end work, sometimes shipping the initial version themselves. Anthropic hit $14B in annualized revenue in February 2026, up from $1B fifteen months ago. Claude Code alone went from zero to $2.5B ARR in nine months. That adoption curve tells you how much value teams are extracting. If your PMs aren't connected to the codebase, they're living in the past.
 
2. Claude Cowork
 
Once you connect Cowork with MCP and API to your data sources (analytics, user research, customer support, CRM) and your output systems (Linear, Jira, your engineering team's coding agent), it becomes the best harness for creating strategy documents, PRDs, and roadmaps. Nothing else compares right now. It works in Google Slides, Keynote, PowerPoint. It builds Excels and impact sizing models. When Anthropic launched Cowork plugins, Thomson Reuters dropped 15.8% in a single day. LegalZoom fell 19.7%. The market is telling you what this tool means for knowledge work. Your team shouldn't be using ChatGPT or Gemini chat. They should have Cowork.
 
3. MCP or API into all key tools
 
People give their teams Code or Cowork access and then don't connect them to systems because of IT constraints, privacy, regulatory, compliance. It totally handicaps the team. If they're not plugged into analytics, business intelligence, actual revenue metrics and not just engagement data, they can't make strategic decisions. The best product leaders are working through the security hurdles and getting their teams connected to everything.
 
Here's your guides:
 
1. Claude Code: https://lnkd.in/gCevZspr
2. Claude Cowork: https://lnkd.in/gwhgmSfw
3. MCP + PM OS: https://lnkd.in/gsH9RSfJ
 
PMs already want to use these tools. The bottleneck is leadership. Work through the IT and security hurdles. The teams that do will be winning in the market.

OpenClaw has 245,000+ GitHub stars and zero real guides for PMs.
 
Sam Altman bought it for over a billion dollars. 2 million people visit it every week. And most PMs I talk to either haven't heard of it or tried the install and gave up.
 
The problem is setup. Terminal commands feel unfamiliar, Slack permissions are confusing, and one wrong config means your bot has unrestricted access to your entire file system. In testing, Naman Pandey's bot read personal files it had no business accessing. Another time it WhatsApp'd every contact in his phone with pairing codes. Including his mom.
 
But once you get past setup, OpenClaw does something no other AI tool can: it acts while you sleep. Set a cron job at 9 a.m. to scan engineering and design channels, summarize blockers, and post a standup brief. Walk into your first meeting already caught up. Set it to monitor competitor websites every 30 minutes. If they change their pricing page at 1 a.m. and overwrite it by morning, OpenClaw already captured it.
 
I spent weeks building the complete setup and PM automation guide with Naman, who walked through every terminal command, every Slack permission, and every security pitfall live on camera.
 
The full breakdown:
 
🎬 Video: https://lnkd.in/g9AYwDef
 
Also on: 
Apple: https://lnkd.in/eCYGNTdG 
Spotify: https://lnkd.in/eyt7agKj
 
What we cover:
 
• Three-command installation (15 min, no coding background needed)
• Full Slack integration walkthrough
• 5 PM automations: knowledge base, standup summaries, competitive intel, VoC reports, smart bug routing by customer tier
• Security audit and deployment options
 
Full written guide: https://lnkd.in/grXq6my6
 
Thanks to our sponsors: Jira Product Discovery · Vanta · Mobbin · Maven · Product Faculty

75% of AI PM interviews are behavioral.

Here's the full breakdown (save this):

I help 100s pursuing AI PM jobs, so I see the inside. And here's the reality of what questions you're going to hear over the course of an AI PM job search.

1. CASE INTERVIEWS (15%)

These get all the attention, but they're mainly in use at the top companies with lots ex-big tech folks like OpenAI, as well as big tech itself.

a. Product Sense (5%)

The AI twist on the classic product sense case is here, and many candidates I have mocked with struggle to adapt without significant practice.

Guide: https://lnkd.in/ekWty_bw

b. Product Design (5%)

The classic 'design a pencil for the blind' has evolved to AI variants like 'design an AI device to communicate with pets', bringing with it its own set of challenges.

Guide: https://lnkd.in/g3gg2pct

c. Success Metrics (5%)

The classic execution interview hasn't left, because this is one of those areas PMs are expected to have a pointy edge compared to their dev counterparts.

Guide: https://lnkd.in/gfffr_zC

2. BEHAVIORAL INTERVIEWS (75%)

a. Leadership & Drive (40%)

This is the entire genre of questions about handling stakeholders and past experiences, by far still the largest bucket of questions. 

Guide: https://lnkd.in/eGhgVNyh

b. Values & Culture (10%)

Every company is their own unique wallflower, and, mostly as startups, they strongly feel that way. Expect to show that you fit that company's philosophy.

Guide: https://lnkd.in/gS-CKH9P

c. AI-Specific Experience (25%)

This is the area every company wants to go deeper beyond the resume: "Did you actually do AI PM work in the past or are you making it up?"

Guide: https://lnkd.in/edKrAAFA

3. TECHNICAL INTERVIEWS (10%)

a. Technical Deep Dive (5%)

These are surprisingly rare. Expect the rare technical question if it's a technical team or role, though.

Guide: https://lnkd.in/dG9EFhiR

b. Strategy / XC Presence (5%)

The PM and designer role are merging. So they want to see how well you can handle it.

Guide: https://lnkd.in/dve8it6D

Be strategic with your practice. 

Want coaching? Apply to the cohort: https://www.landpmjob.com/

📌 Want my free framework to success metrics interviews? Comment 'success metrics framework' + DM me (must do both). I'll reply to those who repost faster.

The AI PM interview is like any test. Break it down into the parts and grade yourself on each. Then bring up your worst grades.

I built this entire deck without ever leaving Claude.

Step by step (swipe through):

Gamma just launched as a Claude Connector.

You can now brainstorm, generate a presentation, and keep refining it all in one conversation. No copy-paste. No rebuilding from scratch.

The combo plays are even crazier:

- Gmail + Gamma: emails become decks
- Notion + Gamma: docs become presentations
- GitHub+ Gamma: repos become project overviews

Your deck finally keeps up with your thinking.

Try it: https://lnkd.in/g7YNKrJY

How to Build Presentations with AI

Brendan Short

Playing long-term games w/ long-term people ✨ Building Groundswell

Currently writing about the future of gtm/tech, at The Signal. Subscribe now (it’s free): www.TheSignal.club 

I’ve worked at a couple of startups ($0-$10M), a couple of big companies (Zoom Video, Apollo.io), and I’ve started a couple of venture-backed SaaS companies (Hexa.ai, Groundswell).

I’ve been an SDR, AE, Head of Sales, SDR/Ops Leader, and CEO. Jack of all trades, master of none, perhaps. 

I like learning and building. 🤘

Writer

• Separating the signal from the noise—a weekly newsletter on all things GTM tech

Brendan Short 🌊

We ArE sO eArLy WiTh Ai

but for real, this is bonkers:

brendan short

Writing The Signal → Exploring AI & the future GTM playbook 🐇 🕳 | Playing long-term games with long-term people 🫡

Salesforce bought Momentum last week. That's 10 acquisitions in 6 months, and nearly all of them are about filling holes in Agentforce.

Momentum is the interesting one. The idea is to take what people say on calls (eg: Zoom and Google Meet) and turn it into data that AI agents can use. Because right now, Agentforce only knows what a rep typed into Salesforce after the call (which, if you've ever managed a sales team, like me, you know less than half of the important information makes it into the CRM).

For the short-term investor narrative, this is great: "We're buying our way to a complete AI-native revenue stack" is exactly the kind of thing that gets an analyst to raise a price target. And it might work! Ten acquisitions that all fit together on a PowerPoint slide is a pretty compelling narrative.

The problem is that customers don't care about investor narratives/slides. They care about how a product will help them win. And the reality is, when a product makes ten acquisitions in six months, that means ten different engineering teams figuring out who reports to who. Ten code bases. Ten sets of gtm leaders jockeying for headcount. Ten product roadmaps that need to become one product roadmap, except nobody agrees on what that one roadmap should look like. The people who are best at navigating that kind of internal complexity are rarely the same people who built the original product.

Salesforce has run this exact experiment before. In 2022 they bought Troops.ai, a Slack-based tool that helped reps update their CRM without leaving Slack. Good product, real users, compelling pitch. A couple of years later, Salesforce killed it, migrated everyone to something called Slack Elevate, and Troops basically ceased to exist. The thesis was right. The execution wasn't.

I don't think this means Salesforce is making a mistake. That said, I do think that the nature of being a platform this large makes it almost impossible to ship fast and stay close to users after absorbing a smaller company. There are simply organizational physics at play. The gravity of a 70,000-person company bends everything toward internal consensus and away from the best interests of the end users.

Which is why I'm even more bullish on my friends over at Attention (Anis Bennaceur). Attention plays in the exact same space as Momentum. Same problem, same category of customer, same bet that conversation data is the context layer that makes AI agents useful. The difference is that Attention doesn't have to spend the next 18 months figuring out how to fit inside Salesforce. They get to keep doing the thing that made them good in the first place: talking to customers, shipping product, and not spending their Mondays in integration planning meetings. Momentum getting acquired actually validates the entire market Attention has been building in.

The question for revenue teams is pretty simple. Do you want a LEGO castle? Or an agile innovator still operating independently?

Writing The Signal (AI x GTM weekly newsletter) | Playing long-term games with long-term people 🫡

taplio

The Battle

Aakash Gupta

Brendan Short 🌊

Followers

Posts

Engagements

Likes

Performance ScoreAakash Gupta's LinkedIn Performance

Performance ScoreBrendan Short 🌊's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Aakash Gupta

Brendan Short 🌊

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Aakash Gupta's LinkedIn Performance

Performance ScoreBrendan Short 🌊's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreAakash Gupta's LinkedIn Performance