We Got into YC, Got Kicked Out, and Fought Our Way Back

This story is being published in the Product Hunt Weekly Newsletter. If you'd like to read more stories like this, subscribe here. 🌟

Hey Product Hunt! I'm Mahyad, the founder of SigmaOS and now Okibi — an agent that builds agents for you, launching today. We're currently going through YC for the second time, but the first time around wasn’t exactly smooth. In fact, we got kicked out… and then fought our way back in. Here's the full story.

Browsers Weren’t Built for Work

In 2021, we started building SigmaOS because modern work had outgrown the browser.

You had Slack, Notion, Figma, Google Docs, and 43 tabs open at all times. Browsers weren’t built for this kind of workflow. We needed something that treated the browser like a workspace, not a filing cabinet.

SigmaOS let you group tabs and web apps into workspaces. Each project had its own space. No more hunting for lost pages or context switching every 30 seconds.

We were proud of it. We thought it was worth applying to YC.

Accidentally Scheduled, Barely Ready

I got the YC interview invite while movers were packing my apartment. I was so excited I immediately called my co-founders, Saurav and Ali.

While clicking around, I accidentally scheduled our interview for two days later. We were completely unprepared.

We split up fast. Saurav worked on the demo. Ali and I walked around London rehearsing our answers, especially the one YC always asks: “What are you building?”

Also, I had made a bet with Saurav. If we got an interview, I’d show up with a mullet and a mustache. So I did.

The First Call

On the Zoom: Gustaf, Harj, and Nicolas. These were people I had watched for years. We pitched. The interview lasted 10 minutes. They told us we’d hear back.

That night, nothing. I messaged the team, joking maybe they were just being polite because it was late in London.

The next day, Gustaf emailed. He liked the team and idea but wanted to see more. He sent a link to book a second interview. I accidentally scheduled that one two days later too.

The Yes Moment

That second call lasted around 30 minutes. The questions were sharper. Our answers were stronger.

At the end, Gustaf asked, “Would you like to do YC?

We said yes and from there on it was “Welcome to YC.” We couldn’t believe it. We screamed. We jumped. We hugged. It felt like years of dreaming and building had finally clicked.

The Launch That Broke Everything

A few days into the batch, we launched SigmaOS on Bookface. Early feedback was strong. The founder of Codecademy signed up, which felt surreal—Codecademy was how I learned to code. Two days later, it all fell apart.

I woke up to messages from Saurav about passports and residencies. I didn’t get it. Then I checked my email. YC had to kick us out.

Because of US export laws, they couldn’t give internal access to people with only Iranian permanent residency. I had one of those. My UK Exceptional Talent visa didn’t count.

The Weirdest Founder Problem Ever

We were not facing a market problem or a growth problem. We were dealing with international compliance.

I told YC they should keep the team and the company, and just drop me. Gustaf said no. He gave us a few weeks to figure it out. If we couldn’t fix it, all of us would be removed.

It was an absurd situation. But Gustaf kept doing office hours with us. That support meant everything.

Code in the Morning, Immigration at Night.

We ran a double shift. Mornings were for building. Nights were for immigration strategy.

Eventually, we found an obscure workaround. Paraguay offered permanent residency if you made a qualifying investment. We found someone who worked in immigration there. She barely spoke English. I barely spoke Spanish. Through WhatsApp and a lot of improv translation, we made it work.

A month later, I became a permanent resident of Paraguay and we sent the documents to YC. Their reply: “This is true resilience.” The legal contact joked, “You’re like cockroaches.” We were officially back in.

The YC No One Talks About

That story became our go-to with investors. But more importantly, it changed how I saw YC.

Yes, YC helps you launch and fundraise. But the real value is quieter. It’s the belief they have in founders. It’s removing roadblocks so you can keep moving forward.

Gustaf didn’t just back the product. He backed us.

The Next Bet

By early 2025, I was winding down my work on SigmaOS. But I had already started playing with AI agents.

We had used GPT-3.5 and GPT-4 in SigmaOS to build agentic workflows. But everything we tried broke.

Building useful agents meant juggling tool calls, syntax trees, hosting, evals, and endless patching. Even our own internal tools eventually collapsed under their own complexity.

Building What We Wanted

That’s when we started Okibi. It’s a web app where you build agents using prompts. No fiddly UIs. No weird wrappers. Just explain what you want and collaborate with it. It writes the code for you, sort of just like Loveable.

We tried all the automation tools like Zapier, Lindy, Gumloop, but they never felt fast enough. We always defaulted to doing things manually. Okibi is for people who want power without setup.

Back Again

As soon as we started, I knew we were applying to YC again.

Not because of the check. Not because of Demo Day. Because YC pushes you to move faster than you think is possible. We set the application deadline as our launch deadline. We shipped on time. We got the interview.

The next day, we got the call. We were in. This time, there were no mullets. Just momentum.

That brings us to today, Okibi is launching on Product Hunt, so if this story caught your eye, and you want to learn more, check out our launch.