From zero to $1M ARR in 36 days by publicly messing with Big Tech

Roy Lee went from broke, unemployed college sophomore to founder of a $1M ARR product in 36 days. Now, Interview Coder is at $110k MRR and growing 20% week over week.

If he can make it work long term, it might even be worth getting blacklisted and potentially expelled.

Here's Roy on how he did it. 👇

Zero to to $1M ARR in 36 days

I'm a sophomore at Columbia University studying computer science, though I might be kicked out by the time you’re reading this.

My only current product is Interview Coder, an invisible, real-time AI assistant for technical interviews that helps candidates pass Leetcode-style coding assessments without getting caught.

I built the core product in 4 days, fine-tuned it through real job interviews at Meta, Capital One, and TikTok, and posted about the experience, going very viral in the process (50M+ views).

By the time I hit $10k MRR, I had already blacklisted myself from big tech, faced disciplinary action at school, and was en route to expulsion. Wasn’t the worst thing in the world, but was certainly pretty nerve-wracking — especially for just $10k.

But within just 36 days, the virality coming from the back-and-forth between me and Amazon, as well as between me and Columbia, catapulted Interview Coder to $1M ARR. We just crossed 110k MRR, growing at about 20% week over week, with about 95% margins.

Coming up with a million-dollar idea

I’ve been a chronic rule breaker since I was born. I spent many weekends at Saturday school and detention during high school, so I was naturally very interested in fucking with the state of technical interviews when I figured out what Leetcode was.

The idea for Interview Coder came when I heard a story of my friend using his phone + ChatGPT to pass a technical interview. Ridiculously stupid and risky, and it would be super easy to build a fix. I was a broke, unemployed college kid and decided to give it a try.

The validation was pretty immediate and we were doing $10k MRR almost as soon as I launched. The nature of what I was doing for the marketing was extremely controversial, so we went pretty viral consistently.

Deep links with Electron are a b*tch

My cofounder, Neel Shanmugam, and I built the core tech of Interview Coder in 4 days, using Electron, React, some serverless edge functions hosted on Vercel, and the OpenAI API. Really simple product.

Took about a month to tweak the UX and paywall it/auth protect it — deep links in Electron are a bitch.

Beyond deep links, my biggest challenge in building was getting DDOS’ed early on. It cost about $2kin Vercel hosting + API credits.

Decisions about the product were entirely made based on my live testing of the tool throughout the recruiting season. I was the only user for the majority of the building process and used live interviews to test new features.

I had the time to do this because I was already skipping class and dicking around coding on stuff anyway. And it was cheap — it cost like $20 a month at the start.

Subscription + credits

We only do subscriptions. We started at $20/mo, then figured out that this was too cheap. So after about four weeks of that, we bumped it to $40/mo and found that pretty much everyone was still paying.

So, after about 4 days of that, we bumped it to $60/month which is where we’re hovering right now.

We also recently added a $300/year annual subscription, but not too many bites on that.

We also charge $10 to refill credits and allow for more uses when someone hits their monthly limit.

Insane viral stunts

I just relied on doing insane marketing stunts to go viral.

The product itself was really built for virality since it was so taboo. So there wasn’t much to do except post what I was doing with it.

The only problem was that I got banned from LinkedIn and Reddit for posting about it, so it was hard to ride that growth curve. It wasn’t until I blew up on X that it really took off — thanks, Elon, for not kicking me off.

Two guiding principles for success

Here are the two guiding principles that I found success with:

• Distribution is king

• Nothing is as risky as it seems

And some advice: All your focus should be on going viral, in my opinion. Once your product works, nothing is more important than a wide top of funnel.

This is important even in the ideation phase. Don’t build a product you don’t think is going to go viral.

What's next?

Probably going to try and scale this to $10M ARR.

And right now is pretty challenging because I have to juggle, like, ten things at once, so I plan to hire someone to help with growth and coding.

You can follow along on X and Instagram. And check out Interview Coder.