From Deebug intern to VALKI co-founder
I was part of Deebug before I knew what I was doing. By 2024 I was co-founding a product inside it. Here's what that arc looked like.
Deebug has been around since 2022. My friend started it, and I've been involved since around then — loosely at first, more seriously over time. By 2024, we were building VALKI together and I was taking on the co-founder and CTO role for the product.
That progression — from "hanging around a friend's company" to "actually responsible for a product with real users" — happened gradually enough that I didn't notice it until I was already in the middle of it.
What the early Deebug days looked like
In the beginning I was mostly just around. Helping where I could, learning more than I was contributing. My friend had founded the company; I was somewhere between collaborator and intern, depending on the week.
That period taught me a lot, mostly because I had no pressure on me yet. I could watch how decisions got made, how technical problems got approached, what it actually looked like to ship and maintain a product — without being the one accountable for any of it.
It's a good way to learn, if you're honest with yourself about the fact that you're learning and not yet doing.
Taking on VALKI in 2024
In 2024, we decided to build VALKI. I came in as co-founder and CTO of the product — which was a different kind of responsibility than anything I'd had before.
The difference wasn't the title. It was the accountability. When something broke, it was my call. When we needed to decide between two technical directions, I owned that decision. When the product grew faster than the architecture could handle, that was also mine to sort out.
That weight is hard to simulate in any other way. You can build a hundred side projects and never really feel it — because there's no one depending on what you ship, no one who loses trust in you if it breaks.
What I took from it
The main thing: ownership changes how you think. Not just about the product, but about every technical decision that feeds into it. You stop asking "will this work" and start asking "will this hold up six months from now when we have ten times the users."
I don't always get that question right. But at least I'm asking it.