proof of work.

I was in my first year of college, living in a hostel with my brother. We didn't have much going on — just classes, hostel life, and a lot of YouTube. One evening I stumbled across a browser extension called Harpa AI. The moment I saw what it could do, something clicked in my head — I couldn't believe this kind of thing existed. I immediately called my brother over to show him. He was blown away too. He asked me the name — but I refused to tell him. I wanted him to find it himself. So he searched, and instead of Harpa, he landed on Merlin AI. That was the first time I ever saw that name. Neither of us knew it would mean anything to me.

Fast forward to August 1st, 2024. I applied to Merlin AI. Formally, through the proper channel, like everyone else. The application expired without a response. That was that. I moved on.

Then on September 10th, 2024 at 3:26 PM — I was walking back from my college classes to my hostel, just a regular afternoon — my phone buzzed. It was a message from Siddhartha sir, the CTO at Foyer Work, reaching out for an interview. Not a recruiter. Not an automated email. The CTO personally. Of the same company that had already let my application expire 40 days earlier. I still cannot fully describe what I felt in that moment. I had been genuinely praying — not just hoping, actually praying — for any kind of work. Freelance, unpaid, coffee money, anything. I just wanted to build something real for someone. And out of nowhere, this message arrived. The interview was the next day. I barely slept. I wasn't confident — I had no idea if I was good enough. But I got in.

From the very first day, writing code for Merlin never felt like work. I don't know how else to say it — it just didn't. While other people clocked out, I was opening my laptop. Not because anyone told me to. Not because there was a deadline breathing down my neck. But because I was in the middle of a problem and leaving it felt physically wrong.

I remember the night I built the AskMerlinAI Twitter bot. It got assigned to me at around 11:30 PM. By the next morning, it was done and live. That night wasn't unusual — it was just one I happen to remember clearly.

There were nights migrating entire codebases from one platform to another, alone, at 3 AM, when the rest of the world was asleep. Nights where I'd start on something small and look up at 5 AM wondering where the time went. I wasn't grinding. I was genuinely absorbed. There's a difference — and I felt it every single time.

My goal from the very beginning was never to just ship things. It was to ship things that actually matter — features that real users would feel, improvements that moved the product forward in ways that counted. Every time I opened a PR, I wasn't thinking about adding to a number. I was thinking about the person on the other side who was going to use what I just built.

I've shipped features I'm genuinely proud of. Deep Research, Thine, Tool Orchestrator, Crafts, Streaks — these aren't items on a list to me. I remember exactly where I was when I first understood the problem, exactly what it felt like when it finally worked, exactly how many times I tore it down and rebuilt it because the first version wasn't good enough yet.

The stats below aren't something you can manufacture. You cannot push 329 PRs across 14 repos in 17 months on obligation alone. Obligation runs out. It runs out at 6 PM, it runs out on weekends, it runs out the moment nobody is watching. What you see below is what's left when you genuinely cannot stop.

These aren't numbers. This is just what makes me happy.