Check it out on Apple, Spotify, or YouTube.
Brought to you by:
AI PM Certification: Get $500 with code AAKASH25
Jira Product Discovery: Plan with purpose, ship with confidence
The AI Evals Course for PMs & Engineers: You get $800 with this link
Today’s Episode
Amazon banned him. Columbia kicked him out. Harvard rescinded his offer.
Now, he’s gone viral repeatedly, built a $6M ARR AI startup in 20 months, and raised $15M from a16z… in weeks.
He is Roy Lee, the founder of Cluely.
Cluely is the AI tool that helps users “cheat” during meetings, interviews, and sales calls - with an invisible overlay and real-time answers.
But the product is only half the story.
Roy also built a viral content machine that is dominating the X timeline.
If you remember my AI Job Search Tool Market post, I rated InterviewCoder as one of the best. This is that founder.
And his goal is to build an AI tech company rivaling xAI and OpenAI.
Here’s what you’ll learn if you listen/watch the whole thing:
1. The life story of a 21-year-old who built an AI product, went viral, and crossed $6M ARR… within weeks
Roy Lee was never built for playing it safe. Even as a kid, he was loud and polarizing.
He’d say whatever was on his mind - in class, online, anywhere.
He racked up detentions, got suspended repeatedly, and eventually, his Harvard acceptance was rescinded after a group of students reported his behavior.
Later, he got kicked out of Columbia too (as I shared back in March.)
Most people would’ve taken that as a sign to lay low. Roy took it as fuel.
It began with a product called Interview Coder
I called it the best sole purpose AI Job Search Tool on the market. He built it.
A tool that helped you cheat on technical interviews using AI. But it wasn’t just another copy-paste chatbot.
Roy built built ‘liquid glass’ before Apple - a transparent code overlay that let you see suggestions without switching tabs or disrupting your flow.
It was new. It was aggressive. And it worked.
He passed the Amazon interview, got the job, all while cheating with his tool and HE RECORDED it all - from start to finish!
Then came the launch moment, a video that straight-up said:
“F*ck Leetcode. You don’t need to grind anymore. Just use this.”
It exploded across tech Twitter, Youtube, TikTok, press, anywhere!
And just like that, Roy was a viral presence. But not without consequences — the backlash got him banned from Reddit, Discord, getting banned from companies & some universities.
Still, the attention stuck.
That momentum turned into something bigger: Cluely.
He didn’t stick to the Interview Coding Tools Market. He decided to go after the broader AI market - even positioning himself in the future against Sam Altman and Elon Musk.
I’m serious. That’s his ambition.
Instead of just helping you cheat interviews, Cluely let you cheat everything:
→ Sales calls
→ Zoom meetings
→ Presentations
→ Exams
→ Daily workflows
It listens to your screen and audio, and gives you contextual help in real time... Surfacing info instantly, invisibly, and undetectably.
2. Raising $15M from a16z without even trying!
Most founders spend months pitching VCs, tweaking decks, and chasing intros.
Roy didn’t even open a pitch deck.
He raised $15 million from a16z in two days… without formally fundraising at all.
Here’s how it happened:
As soon as Roy and the Cluely team landed in San Francisco, word started spreading.
The product was going viral. His content was everywhere.
Investors noticed.
Within days, a16z showed up at their office.
They saw what Cluely was building.
They saw the momentum.
And they moved fast.
Roy didn’t run a process. He didn’t shop the term sheet. He didn’t need to.
That’s the power of being undeniable. When your product is breaking the internet, the checks come to you.
3. Roy didn’t “get lucky” going viral. He built a system
Most founders treat virality like a lottery. Roy treats it like engineering.
Here’s his viral playbook, step by step:
Step 1: Understand the new game
Short-form content has fully replaced long-form in terms of reach.
But the number of good creators? Still low.
That means: “If you post something that deserves to be seen by millions… it will be seen by millions.”
Roy realized there’s a massive gap which was not enough viral content to meet demand.
That gap is the opportunity.
Step 2: Build a team with “viral sense”
Every day, Roy’s team does unstructured brainstorms.
The goal:
→ 20 ideas per session
→ Each idea must have 100M view potential
Everyone involved is an influencer.
They don’t guess. They scroll TikTok, study trends, reverse-engineer hits.
They treat virality like a craft, not a fluke.
Have you ever seen a time announcement post gets… 11k+ likes?
Everyone on their marketing team is a creator with good following, not people with degrees.
Step 3: Optimize for 2 things only
His rule is simple:
A post goes viral when it’s:
Digestible
Reaction-worthy
If it’s not easy to understand at a glance… it dies.
If it doesn’t spark opinions… it dies.
You need content that the average person wants to quote-post, argue with, or share instantly.
That’s what keeps you at the top of the timeline.
His own post got 67.1k+ likes!
Step 4: Forget funnels, track vibes
Most marketers obsess over attribution and conversion metrics.
He ignores almost all of that.
Because the formats that work today… won’t work next week.
So instead of trying to squeeze ROI out of every video, he focuses on visibility.
“If I saw this video as a consumer… would I be interested in the product? That’s the only test that matters.”
Step 5: Make stunts part of the roadmap
Some startups plan sprints.
This team has planned some of the viral stunts that got media coverage and went viral with just few thousands of cost.
→ Stripper ads
→ Parties that get shut down by SF police
→ Launching a full-blown anime trailer for Cluely
None of it is random.
Each one is designed to hit the emotion + distribution combo that gets people sharing.
And it works.
If you’ve read this far, I’m sure you’ll LOVE this episode :)
Key Takeaways
Related
Podcasts:
LogRocket CEO and Founder on How to Build a $100M+ AI Startup
Amplitude CEO and Founder on Building the Product Analytics Leader
Newsletters:
If you prefer to only get newsletter emails, unsubscribe from podcast emails here.
If you want to advertise, email productgrowthppp at gmail.
P.S. More than 85% of you aren't subscribed yet. If you can subscribe on YouTube, follow on Apple & Spotify, my commitment to you is that we'll continue making this content better.
Share this post