Jay Singh

CEO @ Casper Studios | Ex-Linkedin: We help companies adopt AI and build AI-powered products.

I am a co-founder of Casper Studios, a product studio that develops and integrates AI solutions for companies. I previously was a business development and strategic partnership manager at LinkedIn. With over eight years of experience in these fields, I have a track record of creating value for customers, partners, and stakeholders.

At LinkedIn, I have helped lead the global expansion of identity verification capabilities to key markets, such as India, Europe, and LATAM, enabling over 80 million members to access trusted and secure services. I have also driven 20+ partnerships with leading companies and governments, such as Clear, DigiLocker, and Hyperverge, to enhance the user experience and the network effect of the verification project . 

At Casper, I have leveraged my expertise in AI and product development to create a popular chrome extension that assists users in content creation and summarization, attracting over 20,000 users in less than six months. I have also generated $x00,000s of revenue by offering tailored strategy consulting, product development, and design services to other businesses that want to integrate and leverage AI.

I am passionate about making products accessible, efficient, and user-friendly, and empowering people and businesses to unlock their full potential through technology. I have a strong background in trust, verification, and HR tech, and a keen interest in entrepreneurship and innovation. Outside of my work, I invest time in supporting the AAPI community, love meditating, and dancing with my friends.

Co-Founder & CEO

With a team of experts from LinkedIn and Microsoft, we help companies integrate AI into their workflows, from strategy and prototyping to complete product development.

Schedule a call with us to see how we can help.

Great article from Gergely Orosz. Below is an excerpt from a convo with the CTO of Atlassian on how he had to go buy his own laptop just to play around with new AI tools!

That’s kind of wild.

If even senior leaders have to jump through hoops from their own IT teams to experiment, something’s off. Big companies need a different procurement mode otherwise they won’t move fast enough to adopt these tools.

Jay Singh 

CEO @ Casper Studios 👻 | Ex-Linkedin: We help companies implement AI

Built a fun automation that I think will help us close an additional $5-10m in revenue over the next year or two... Here's what we did in in a few hours:
1.) Exported my LinkedIn connections
2.) Uploaded it into Cursor with Claude Code
3.) Asked to filter those folks by our ICP in Cursor
4.) Ran each person and company through a research query (who they are, what problems they're solving, what their peers are doing in AI)
5.) Connected that it to our Google Drive case studies + Fireflies meeting transcripts via MCP

The output? A personalized note for each person....

"Hey Jay, looks like we've been connected for a while. We've already helped companies like yours with [specific use case]. Here are a couple case studies. Happy to trade notes on AI if you'd like :)"

I'll likely still massage the messaging. But this is so powerful.

Max is one of our tech leads. He’s also become a bit of a celebrity in the San Diego AI scene 💫 

He’s been hosting a number of evente, some directly in partnership with Anthropic. 

The early vibes and format that Max and the team experimented with is now getting formalized and rolled out more broadly under the name Claude Impact Labs, which Max will be hosting its first one next Saturday. 

If you’re in SD comment below and I’ll share the link + connect you with Max :)

We spend so much time chasing strangers for sales. I think there's millions in revenue through the people we know. We built an agent that runs our LinkedIn connections against our ICP and bubbles up the ones that matter.

When we posted about this last week we had so much interest that we had our intern, Shaan, turn it into a simple tool you can use.

How it works:
1.) You land on the homepage. It walks you to LinkedIn to export your connections (takes ~10–20 mins).

2.) Upload the CSV back into the tool.

3.) Define your ICP. I usually just dictate it — “NYC fund managers, $1B–$10B AUM,” etc.

4.) It runs through an LLM and surfaces the people in your network who match.

5.) You can click in, research deeper, generate a message, and reach out.

That’s it. It’s free for now. We’re basically burning through Shaan’s monthly token allocation and seeing what breaks lol.

If this would be useful, comment LinkedIn ICP and I’ll send it over. We’ll probably cap it at 10ish people while we test this out!

You can be at a romantic dinner in SF and the table next to you is literally debating Opus 4.6 vs Codex 5.3. This city is a meme. Full of nerds. I love it.

Yesterday I spoke at the The AI Lab Community meeting with Ryan Kurt and Alex Lieberman on how to implement AI into enterprises for GTM / sales.

One thing I walked through: how we have automated our own entire go-to-market flow at Casper Studios. Works across three key steps: 1.) finding a lead, 2.) closing a lead, and 3.) starting the client engagement. 

None of this required a dev team (but may be faster if you involve them). Just Claude and your existing data. Let me know if you have any questions - happy to walk you through any part of this.

Comment below if you have any q's and I'll answer them!

Our team just made some big upgrades to the site. We added 5+ new client case studies, tightened up how we describe our offerings, and secured casperstudios.com instead of the .xyz.

It feels like a real step forward :)

The main insight on how we already working with large F1000s? We're getting procured through their consulting firm of record; the one that’s been working with them for 10+ years. We join their team, and they deploy us to their larger clients.

This helps in several ways. We can move faster to provide value to clients. We can also provide value to the consulting or agency of record by teaching them how to use AI faster. And then when the trust is high, we can cross-sell into other clients they have in their portfolio.

This is where the BD and partnership chops really come in handy. It’s also why I don’t really think, at least today, anyone is a “competitor.” Show me a competitor and I’ll tell you how we can support each other and grow both of our businesses. 

Our GTM strategy for the next 2-3 years is literally to be best friends with everyone.

“I feel like everyone’s working on Saturdays now.” – me. Ria’s response: “I think that’s because everyone’s having more fun with their work now.”

She’s right. Texts, Slack messages, emails, LinkedIn DMs, they don’t stop on weekends anymore. The dialogue is more expansive than during the week. 

It can be intense. Take the breaks when you need them.

But 90% of the time? It’s just super exciting.

Was speaking with a PE friend who's exploring buying a traditional SaaS business for $5m-$10m with plans to "AI-ify" it. My honest (and tbh biased) take: just rebuild it from scratch with AI instead.

Work with that team, understand the workflows they offered to clients deeply, and give them a piece of the company to go on that ride with you. 

It's a harder sell emotionally (but what about all the work I've done over the past 5+ yrs...), but retrofitting AI onto an existing business is slower than starting from first principles and asking "what would this look like if AI was at the core from day one?

We help companies with both, and honestly enjoy both. But if you have a choice, and the business is still relatively small (< $10m ARR), only SaaS oriented (I think services is a different for a few reasons I’ll unpack later) the difference in velocity and change management effort is so significant that starting from scratch might just be the smarter move.

**Editing Post With More Nuance:
Had a nice sleep, reflected on this, and wanted to share a few more thoughts.

The nuance I would include is this: if this buyer is able to bring alongside some members of team that built the product, including several key relationships with the clients with them - esp the ones looking for AI-first capabilities (some may even be exploring AI first competitors), then it becomes more relevant to do. 

My POV on this is being driven by a few of our own clients. 

Ie they were leaders in a company in a certain space. Some of them have made the decision to step out, rebuild, bring a few of the clients with them, and finding success. There’s legal hurdles depending on the industry you’d have to be wary of here too. 

Both can be successful here! I think for even the companies <$10m that want to implement AI I’m pushing them to think about ways to rebuild from scratch in some corner of the company, to experiment, unencumbered by the procurement process of onboarding tools, by older ways of thinking, then try to push those features after into the product. 

Happy to be wrong on this too :)

We're hosting a small (10-15 friends max) dinner for AI BD / GTM / Partnerships leaders in SF next Wednesday Feb. 18th. Want to come? Comment below or DM me. 

TL;DR on the idea:
-Share what's actually working in AI go-to-market right now
- Talk about what's not (equally valuable)
- We're a few weeks out from testing our new GTM engine and would love to share what we've learned building it

Dinner's on us. No decks. No sales pitches. Good vibes.

Who should be there? Tag someone below or DM me if you're interested. Will be selective to specific BD/GTM who are at AI companies this time - and hope to expand this in future dinners! 

I’ll share the link in the comments shortly too.

Cc Martijn Lancee

Companies are still underestimating LinkedIn for sales and talent acquisition. At Casper Studios, we’ve had 10+ teammates to create content. The velocity of new and repeat business is increasing. Higher-quality talent keeps reaching out.

All content is high quality. It has to clear our bar. Beyond that, the team has real freedom in how they show up and what they talk about. It works way better when it actually feels like them.

Behind the scenes, we have agents running that:
-Extract insights from our meeting transcripts
-Monitor competitive research to see what people are talking about
-Flag risks, tighten language, and make small improvements (we still miss things like we did today on one, but it gets better every week)
-Bonus: As things get more complex, you can use these to also reach to our audience. We’ve been leaning into that more recently

None of this works if the founder or CEO isn’t leading. You can’t just tell your team to “do AI” and expect it to happen. You can’t just tell your team to “do content” and expect it to happen.

Those are now core parts of the job.

Once you start, something interesting happens. The team has fun with it. They get inbound for their time. They meet sharp people. They share what they’re building for on platform and internally.

And once that starts, you feel the pull of the platform and it's addictive. 

If you want the suite of agents powering this, comment below. If there’s enough interest, we’ll work to externalize them

I'm hesitant about using AI for writing. Although, this pipeline has been helping Martijn, Derek, Shaan, Leonard, Sebastian and others on our team create a lot of authentic content lately. Here's our secret sauce...

The tl;dr is that the data sources matter. 

Ours are typically Fireflies transcripts, Slack messages (specific internal channels with sources our team shares), Google Docs and Drive. 

We run all of that through a processing layer with different skill files to generate ideas and refine them. Most importantly, it scrubs all client-specific information. Everything is hosted on our internal environment. Then we review and post.

Did I write this post myself without AI? Yup. Do I think AI, connected to your internal data, can help you generate great content too? Yes.

Shoutout Shaan Jain for coming in super hot to the Casper team in the last few weeks and building this out. 

If you want us to set this up for your company comment "ugh why are you making me comment for this" below hehe.

Been working with Zachary Jacobs to build the first version of an AI voice product in the GLP-1 space. Feedback like this is what makes this fun!

I’m also enjoying the range of clients we’re working with right now - $10B+ revenue healthcare companies, $2B AUM hedge funds, but also founders.

I think the cross-pollination across those segments is where a lot of the edge for Casper Studios comes from. Great work Elizabeth, Sean, Amiel and Oscar on this :)

Across the board, last week was the busiest I’ve seen for AI services partners in the last few years. Inbound was up everywhere.

Something seemed to flip. Maybe public market noise? Maybe board pressure? Hard to pinpoint.

But everyone is locked in right now. It’s inspiring to observe in our own company and our partners companies. 

Really does feel like a once in a lifetime moment. We’re excited to be in the middle of it.

We were on a sales call with a $1b company and after our intro they said, “So you’re basically the Best Buy Geek Squad for AI?” Hilarious.

One of our clients raised ~$500m last quarter bringing them to $2B AUM. Performance was a big factor. Their LPs also told them the AI platform we've been building together is "the most integrated use of AI from a manager we've seen to date."

Now we're hearing the same thing across the board with other GPs.

A private credit fund with $500m AUM told us their LPs are pushing hard for AI adoption. Their reason was because every dollar spent on headcount is a dollar off net returns. 

A PE fund manager said this week that LPs want to see an AI efficiency roadmap before committing more capital. 

In 2025, AI was offensive, a way to differentiate and raise more capital. In 2026, it's becoming defensive. LPs aren't impressed by AI anymore. They're expecting it.

It makes sense. When LPs evaluate managers, they're buying trust in the sophistication of your process alongside your ability to generate returns. I think AI adoption has become a signal for that sophistication.

Excited to be at the AI in Healthcare Summit in SF this Tuesday. Appreciate the invite Chris Taylor and the Fractional AI team. 

Both Chris and the team have been amazing supporters of myself and Casper over the last few years. It’s been incredible to see their growth. 

If you’re in town for the event and attending - let’s connect :)

No alternative text description for this image

I’ve been noodling on a piece of short fiction set in SF, 2035. Sharing early thoughts - curious to get your feedback….

The premise: What happens when we finally learn to quiet the voice in our heads (the anxious one, the 3am one, the one Buddhism, meditation and therapy spent millennia trying to quiet) and then replace it with a “better” AI one?

The story follows a guy named Jay (lol) through a single Tuesday. Morning coffee chosen for him based on his cortisol levels and sleep scores. Work meetings where both sides are being coached in real time by their AI. A bike ride over the Golden Gate where the product schedules 90 seconds of silence at the exact moment the ocean view hits. Because it knows that view will move him.

The tension is a mix of utopia and dystopia. That’s what makes this and life generally more interesting to me; especially when our world is so black or white. 

The AI makes his life better by every measurable standard. He’s faster, healthier, better at his job. 

The issue is that he can’t remember the last time he made a decision without it. He doesn’t know if he likes his own coffee. He can’t choose a cereal. The muscle for deciding anything has gone away.

Then his AI goes quiet for an afternoon. There’s a bug. And the world gets louder and messier and harder and more alive than it’s been in years.

He tells his wife. He tells his friends. “I’m feeling naked”. He tries to convince others that they should try turning the voice off too. It’ll be more authentic. Some listen others think he’s a Luddite. 

Sharing this early because I want to pressure test the concept. Does this resonate? What’s missing? Any other themes I should pull on?

Current travel run this month: Costa Rica (for fun) → Indianapolis (work) → Des Moines (work) → Montreal (work) → back to SF. All in  ~2 weeks. Derek and I are clearly on tour at this point. Where should we go next??

James Raybould

I've recently joined Turing, an AGI infrastructure company, to lead our Turing Intelligence business, building powerful AI systems to answer the question, "How will AI transform my business?" We combine expert human and artificial capabilities to accelerate companies.

I'm a Generalist. After Consulting at Bain pre- and post-HBS, I joined LinkedIn in early 2010 and spent 12+ years moving across functions and leadership roles: I started in Sales Strategy and Analytics before moving to Marketing, Product Management, and, finally, General Management. I then moved to fractional exec roles at Speak, WisdomAI, and StealthAICo, blending product, GTM, and GM skills; I like breadth above all. 

To tell my longer story, 3 dates from 3 decades help:

Sept. 22, 1995: I travelled 5,000 miles from my London hometown to Stanford. I was drawn to studies, sports, sunshine, and a Bay Area culture more focused on creating the future than curating the past. I rowed on the Stanford Varsity Crew team and studied psychology for my B.A. and M.A. And I've stayed close: I can see the Stanford campus from our backyard.

Sept. 13, 2004: I joined Bain Bay Area as an entry-level Consultant. After a few years of racing triathlon and working in sports business, I was drawn to Bain’s entrepreneurial culture and fast-paced learning environment. Looking back, my strategy and comms chops were forged at Bain, as was a lifelong love of Excel, PowerPoint, and Cinnabon.  

March 8, 2010: I left Bain to join a pre-IPO LinkedIn with ~500 employees and ~$100M in revenue. I discovered a vision that deeply resonated - to create economic opportunity for every member of the global workforce - and, second, a set of inspiring leaders teaching me how a culture and values focus could underpin incredible businesses. Over the next 12.5 years, my roles and responsibilities varied: in function from sales strategy, analytics & insights to marketing to product to GMing; in scope from being an IC to leading ~100 person teams; and in scale from helping run our largest, multi-billion dollar recruiting business to jumping in to help lead our acquisitions of Lynda.com (2015) and Glint (2018). 

Outside of work, I'm a long-term elite endurance athlete (rowing > Ironman > marathons > bike racing > marathon swimming) and compulsive reader and film-goer. Here's ALL my books & films (bit.ly/z2D1sX) and I publish a "What's the So What?" newsletter distilling favourite books and experiences into what matters most (bit.ly/3q188jL)

Founding Member & Advisor

I’m excited to be a founding member of Round, a newly launched community for tech’s most thoughtful leaders. Since joining I've benefited in a major way from small "core group" meetings, special events with luminaries, connections with other tech leaders, and more. Starting in early 2023 I've also begun to help the Round leadership with a range of projects to help them grow their impact.

6 AI agents are working for me right now. In parallel. Like 6 direct reports — but faster, more patient, and available at 7:15 pm on a Thursday.

That screenshot is my monitor tonight. Six Claude Code windows running simultaneously — reformatting a 30-page document, deploying a "kiddo points" web app for my kids, analyzing my health data, crunching LinkedIn analytics, cataloguing every film I've watched, and researching a VC firm's team.

To bring the last one to life, in <60 minutes of clock time and <10 minutes of my actual supervisory time, Claude Code helped me to do some in-depth research on the 60+ person investing team at Bessemer Venture Partners:

First, it pulled all the bios (easy for humans, but would take >60 minutes of mindless copy-and-pasting)

Then, it created the So Whats of their team composition across educational backgrounds, prior employers, and uncovering Quirks and Hidden Patterns: https://lnkd.in/dJjqKqQh

I’ll bet even folks who work at BVP - hello, Shannon Brayton, Lauri J. Moore, Steve Kraus, Kent Bennett, and more - will learn new nuggets from the Quirks section at the bottom of the Notion doc (BTW, the Claude Code-Notion integration is incredible)

This "So Whats" analysis would have taken me 5+ hours to replicate, and, realistically, the quality of my work would have been far lower, and I would never have invested the time to begin with.

And that was the output from just one of six windows running at the same time.

6 agents in parallel probably sounds crazy to many of you.

But with the current pace of AI innovation from Claude Code/Cowork and others, I'm confident it'll be commonplace for everyone reading this by June.

And 6 parallel windows will soon become 60. And then 600...

The bottleneck isn’t technical skills anymore. It’s the ability to generate ideas worth building and to use ever-improving AI tools. 

If you're reading everything about Claude Code — as of today, with Opus 4.6 now available! — but not yet using it, you're missing out on the single most powerful product ever built.

I've never entered a Computer Science building and can't write a line of code on my own. Now I ship new analysis and new products every single day, with zero other humans involved. #JFDI

Inspired by my time at LinkedIn, I built a "Relationships Matter" app that shows which of my relationships are heating up or cooling down. Spanning personal and professional, and all my digital channels.

If you've worked at LinkedIn, you'll know "Relationships Matter" well; it's one of the company's foundational values from day one. But actually keeping track of relationships has always required too much manual effort. 

AI advancements and API integrations have changed that.

My app connects to my full digital life: iMessage & WhatsApp, email, phone & FaceTime, LinkedIn messaging and commenting, and my Calendar, including video versus in-person. 

I had fun geeking out on the scoring and features:

- 📞 It weighs the value of different interaction types — an in-person cup of tea vs a FaceTime call vs a quick text vs a 👍 on WhatsApp
- 📉 Diminishing returns on volume — 50 texts in a day isn't 25x more meaningful than 2
- ⏳ Recent interactions count more — that SF stroll last week matters more than a long catch-up two years ago
- ❤️ Personal relationships need more regular contact than professional ones, since I want family and friends to come first
- 🔄 It tracks who's reaching out more — am I more often the initiator or the initiatee?

I’ve no idea if the scoring is right, but it's a fun one to noodle on and super easy to change as I tune it. 

Every person gets a Warmth score and accompanying tier (from Inner Circle down to Dormant) plus a trend (heating up, cooling, or stable).

I've also set targets for folks — whom do I want to be closer to — and the app shows me how I'm tracking and makes suggestions.

And, NO, I won't be sharing my target tiers or what rank you are on my Relationships Matter list, beyond stating that, YES, my wife is the clear #1 ;)

I built this last weekend in <1 day with Claude Code. A year ago, I would never have even considered tackling this. The barrier to building whatever we want is getting lower and lower.

Still very much WIP. What would you want in a “Relationships Matter” tool?

And if we haven’t caught up recently, consider this your nudge. Reach out or expect a ping ;)

#RelationshipsMatter #JFDI

Traders assume a $30K+/yr Bloomberg Terminal. Recruiters assume a $10K+/yr LinkedIn Recruiter seat. AI tools are next. And the bill will dwarf both.

It's starting with engineers. They're already turning down offers over which AI tools they get: Claude Code and Cursor are becoming a requirement — teams still stuck on Copilot are losing candidates. And token budgets north of $100K/year per engineer are coming fast.

And this won't stay in engineering. As every role builds with AI - PMs, marketers, ops leads, etc., more and more roles will hold the same assumptions.

This changes compensation conversations:
- For individuals: "What's my token budget?" will mean total compensation is salary + bonus + equity + tokens.
- For leaders: we'll need to manage token budgets across teams and roles.

Right now, AI tool spend per employee is perhaps 1-2% of headcount costs. That's heading to 10%. Then 20%. I wonder when the first truly AI-forward companies at scale will cross 50%?

I built a LinkedIn profile analyzer app and accidentally discovered that AI can make people feel deeply understood.

On Thursday, I launched "Profiler," a LinkedIn profile analyzer inspired by Miranda Priestley. I designed it to be accurate and instructive but deliberately snarky and judgmental.

Celebrity profiles were a hit (Stephen Curry is "Formidable (6/7)", Tom Brady only "Flickering (4/7)"), and there was friendly competition — only 12 people out of ~2K profiles analyzed so far are "Unforgettable" (7/7). 

But a surprising number of responses highlighted something I didn't design for. People quoted back the lines where AI read their career and landed on something unexpectedly personal:

🌟 "This phrase may be the best compliment I've ever gotten about my work." - Emilia Varshavsky Shapiro 
😊 "I've had a rough year... and the little quote at the top just made my day." — Eileen Lawless
🙏 "I feel so seen. I guess I've done a pretty good job of intentionally ignoring all the cries and rules around personal branding and stayed true to sharing my open heart and mind." — Todd Lieman
🥲 "Tears. Literal tears." — Lara Frankovitch

If we've been lucky, we've had managers or mentors who said something about our careers that connected dots we hadn't connected ourselves and reflected back something we didn't know was visible. That's a rare gift from a person who knows us well.

An AI tool that this non-coder built over a weekend delivered some of that special feeling in under 20 seconds using nothing but a LinkedIn profile — to complete strangers. I don't know any of the people quoted above.

There's a general perception that AI will surpass us on IQ, but EQ is where humans will maintain our advantage. My weekend project just challenged that assumption. Now imagine what happens when this kind of capability is built with intention — with your full work history, your communication patterns, your actual performance data.

If you haven't tried Profiler yet - it's Desktop only - what it sees might surprise you: https://lnkd.in/drcdRyDc

#AIEmpath #EndBoringProfiles

I'm finding AI is even more about mindset than skillset. 

The more I use it — really use it, getting close to the metal with Claude Code — the more opportunities I see. 

Not because I've learned some special technique. Because my brain is starting to work differently as I see everything through an AI lens. 

More and more, my default when I see friction in my day is, “I could build something to make this better.” Not "find a tool” or "file a ticket”, but to just build it myself. Often in an hour or an afternoon. 

This mindset came only from doing, not from reading.

Which is the opposite of what I’m seeing with too many of my exec friends and colleagues. I see them appointing an AI lead and waiting on a roadmap. Asking their teams to “explore use cases” while they “guide” from above.

You can’t delegate a mindset shift. 

Your team can build workflows and pilot tools all day, but if you haven't personally felt that moment where a week of work compresses into an hour, you're making decisions about the future of your org without understanding the shift that actually matters. 

It's less about being technical than about rebooting your defaults. Stop accepting every process you inherited and start questioning the friction you've treated as normal. 

Our best leaders won't be the ones who authored the best AI strategy decks. They'll be the ones who got uncomfortable early and started tinkering. The ones who yelled at Claude Code, “WTF did you just do?” (while knowing it was almost always because their prompt wasn’t clear enough). 

This mindset only comes from doing. #JFDI

Three years ago, Nvidia's revenue was flat. Literally flat. $26.9B in FY22, $27.0B in FY23. A rounding error of growth. Then this happened:

- FY24: $60.9B (+126%) 
- FY25: $130.5B (+114%) 
- FY26: $215.9B (+65%)

That's +$189 billion of new annual revenue in three years. An 8x increase. A 100% CAGR (Compound Annual Growth Rate) off a completely flat base.

The former Bain consultant in me got nerdy about this, so I went looking for comparisons. I pulled the best 3-year revenue stretch I could find across the major tech companies:

1. Amazon (2018-2021): $233B to $470B; +$237B, 26% CAGR 
2. Amazon (2022-2025): $514B to $717B; + $203B, 12% CAGR
3. Nvidia (FY23-FY26): $27B to $216B; +$189B, 100% CAGR
4. Apple (FY19-FY22): $260B to $394B; +$134B, 15% CAGR
5. Alphabet (2018-2021): $137B to $258B; +$121B, 23% CAGR 
6. Alphabet (2022-2025): $283B to $403B; +$120B, 12% CAGR
7. Apple (FY09-FY12): $43B to $157B; +$114B, 54% CAGR 

Most folks consider Apple's iPhone + iPad era (#7 above) to be tech's defining growth story, but that was +$114B and 54% CAGR. Nvidia nearly doubled it in both absolute dollars and growth rates. WTF. 

One level nerdier: Nvidia has grown by $85B in the past year (+65%). This is #3 all-time for absolute growth: Amazon put up +$106B (+38%) from 2019 to 2020 (peak COVID) off a $281B base, and Apple put up +$91B (+33%) from FY20 to FY21 off a $275B base. 

And going all-in nerd: Nvidia's +$189B revenue growth in 3 years was larger than today's total revenue for each of JPMorgan Chase (founded in 1799), Ford (founded in 1903) and GM (founded in 1908) ;)

You can debate whether AI is overhyped. You can't debate 8x growth in 3 years. 

#WhatWouldJensenDo

Saturday morning with the kids may be the best AI training program available. 

I built "Kiddo Points" for our 8- and 11-year-old munchkins — see screenshots and the view-only version here: https://lnkd.in/dHvVtjuM 

It rewards behaviours across learning, sports, chores, and more. 

And these points unlock smaller choices like picking the restaurant or playing DJ on family drives, and bigger choices like a “double sleepover” with friends or where we travel to on holiday. 

Claude Code made the technical building easy: I ask. It builds. 

The hard part — as with many products — was designing the incentives:
- I don't want them reading - an activity they both love - just to earn points.
- How many tennis serves practiced equates to vacuuming the house?
- Which activities to auto-approve (presume trust) versus require our approval (but verify)?

While building this out, I had fun indulging my marketing geek and creating a promotions system: I can increase points on certain days and for certain activities. I’ll definitely weave in the Winter Olympics in some way; maybe the Super Bowl tomorrow, too.

And I included personal finance education: kiddos can put their points in the bank (and earn interest) or take out a loan (with interest to pay back)!

A parenting app may seem out of place on LinkedIn. 

But building with AI in your personal life is the best side door to learning AI.

No IT approvals. No security reviews. No stakeholders. Just idea → build → iterate. The loop that takes 6 weeks at work takes 6 hours at home. And the AI skills transfer to work.

In addition, spending this much time thinking about how to manage my kids has me thinking about what applies to managing at work.

The problems are nearly identical, although the reactions may be different: my 8-year-old threatens to withdraw hugs if I don't change an incentive, and my 11-year-old already games our system better than any VP I've worked with ;) 

If you're reading about AI but not yet building, make your family the reason to get rolling this weekend. 

#JFDI

"How did you actually build that?" Here’s 1,521 Claude Code messages to answer the most common question I get these days.

Broken into 6 stages, I've pasted every exchange verbatim with Claude (Stage 1: Define and Stage 6: Launch) and Claude Code (Stages 2-5) to build Profiler, a LinkedIn profile analyzer tool:

- The full transcript: https://lnkd.in/dqhUdy9m

In the spirit of "Show, Don't Tell," I hope this helps show the reality of building software with AI in February 2026. A taste of what's inside:

🎭 Defining the voice: "Make sure that it's as though Miranda Priestley gave up running Runway and decided that she wanted to run a LinkedIn profile makeover boutique."

🧪 Keeping Claude code on its toes: "If it failed last time, are you confident this will fix the problem?" Claude Code: "Honestly — no, I'm not fully confident."

😊 Open, honest, and complimentary: “It looks great, nice job.” … “I love it!” … “I think you're right.”

😤 Open, honest, and less complimentary: "Your advice to post it and see what it looks like is absolutely terrible"

🤬 Open, honest, and furious: “I'm IRATE. The bug happened again.” … “WTF”

Overall, building with AI constantly cycles between magic and maddening. From “I can’t believe that 1-hour human task took only 10 seconds with AI” to “My 11-year-old could tell you that makes zero sense.” 

That’s our reality in early 2026, and I hope sharing the full, unfiltered version is helpful for those dipping their toes into the AI building waters.

Tomorrow: why the skills execs already have are their biggest advantage when building with AI.

#ShowDontTell #JFDI

Everyone says AI requires reskilling. They're only partially right.

The skills that matter most when building with AI aren't technical. They're precisely the skills execs have spent their entire careers developing:

1️⃣ - DEFINE the goal
2️⃣ - SCOPE the boundaries
3️⃣ - PROBLEM-SOLVE when stuff breaks
4️⃣ - REFINE with judgment
5️⃣ - MARKET & SELL from day one

Execs aren’t behind on AI. We’re ahead. We've spent decades building the judgment to manage teams; whether you're directing people or AI, the job is the same.

Same skills, different audience.

The barrier isn't technical ability. It's the belief that there is one. You don't need new skills. You just need to start building.

Full article below. #3 in my ongoing “AI for Execs” series.

#AIForExecs #JFDI

The skills that make execs effective leaders are exactly the same skills needed to build with AI. Execs aren't behind — we're ahead. To show, not tell

Open article: AI For Execs #3: Why Execs Are Ahead, Not Behind by James Raybould • 10 min read

AI For Execs #3: Why Execs Are Ahead, Not Behind

In my triathlon days, the people who got fast weren't the ones who prepared the most. They were the ones who sweat the most. AI is no different.

My LinkedIn feed is littered with AI tips, tricks, and debates about whether to use Replit or Lovable. People preparing to build. Not enough folks sharing what they've built.

Worse, all that preparation isn't neutral. It creates a false sense of progress as you feel like you’re getting better at AI, while in reality, the builders are pulling away from you. 

I've seen this same pattern play out three times before:
1. 🏃 Triathlon: So many bought the gear, read the plans, watched the videos, debated the shoes. Did everything except regularly sweat. And then wondered why they hobbled most of the marathon.

2. 📊 Excel: Most of the Excel experts reading this did a stint in banking or consulting. The 17th VLOOKUP or v21 of the model is where it became muscle memory. We don't need to pull all-nighters to learn to build with AI, but it's a whole lot more fun than deal models and decks ;)

3. 👶 Parenting: the jury is still out on whether our 8- and 11-year-old kiddos become upstanding citizens, but my wife and I didn't learn from YouTube or books. We just showed up every day. Imperfectly but constantly.

And if you're not into endurance sports, Excel, OR parenting — why are you seeing my LinkedIn updates? :)

We learn by doing, not by preparing to do. The gap between "I should use AI to build something" and "I built something" is primarily a sweat gap.

I've learned more from weeks of hands-on building than years of reading.

So stop scrolling and fire up Claude Code this weekend alongside me.
Get sweat on your keyboard.
Even better, embrace Wispr Flow and skip the keyboard altogether.

#JFDI

For a non-coder building solo, AI is already so good that it's almost gotten out of the way. What's left is the part that's always been hard: figuring out what to build and why.

I've shipped ~10 projects over the past few months. So far, the tech hasn’t been the bottleneck on any of them. 5 recent examples:

🔍 Profiler: https://lnkd.in/drcdRyDc — a LinkedIn profile evaluator with the personality of Miranda Priestley. The hard part wasn't anything technical, it was making a fictional character funny and helpful at the same time. That's a voice-and-tone challenge.

🎬 My J-Raytings site scoring every film I’ve seen: https://lnkd.in/drNjqRyz (plus /books and /shows). The hard part was design restraint. With 1,700+ films, there’s endless metadata to show, so the challenge was figuring out where more became less. That’s a UX challenge.

⭐ Kiddo Points — an app for my 8 and 11-year-olds to reward a range of habits. The hard part was making sure the points don't become the point — you want kids doing the right thing because they want to, not because they're chasing a score. That's an incentives challenge.

🤝 Relationships Matter — an app to track and strengthen personal relationships. The hard part was building a scoring model for something as complex as friendship. How do you compare a 10-minute text thread yesterday to a 60-minute lunch 9 months ago — and which one matters more for the relationship? That’s a scoring challenge.

🌍 Finally, I advised our 11-year-old on building a geography app: https://lnkd.in/dWCkYQWU.  His problem? How to teach kids about countries and their culture and characteristics, not just quiz them on where they are on a map. That’s a content challenge.

Every example above is a product decision, not a technical decision. We still need designers and engineers to build for scale, but the barrier to shipping something solo is now almost non-existent.

Last night, I built a working AI tutor for my kids in 90 minutes. They’re my beta testers this weekend. Again, the hard part isn’t the tech stuff. It’s deciding how to blend learning and testing and ensuring the tutor’s tone is engaging. Spoiler alert: you can choose from 8 tutor styles, including the perpetual positivity of Leslie Knope from Parks & Rec and the melodramatic menace of General Zod from Superman II.

Yes, there’s a learning curve with tools like Claude Code to do “scary” things like install Python and Node.js or figure out when to use Vercel and Supabase, but 95% of what I’m finding hard are the same product decisions I wrestled with as a PM for years.

The key difference is that waiting months or quarters to ship has become minutes or hours.

If you’re a PM and you haven’t made Claude Code your BFF yet, get rolling. It’ll remind you why you became a PM in the first place.

And if you’re not a PM, it’s the same deal because AI doesn’t care about your job title. It cares whether you know what problem you’re solving.

#JFDI

taplio

The Battle

Jay Singh

James Raybould

Followers

Posts

Engagements

Likes

Performance ScoreJay Singh's LinkedIn Performance

Performance ScoreJames Raybould's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Jay Singh

James Raybould

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);}Jay Singh's LinkedIn Performance

Performance ScoreJames Raybould's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreJay Singh's LinkedIn Performance