I joined Orbit this week.
Before that, I knew the shape of the job I wanted. A small team, real ownership over the work. Not a rough idea, a specific one, close to a checklist. So I searched for exactly that.
It produced almost nothing. Roles that matched the description on paper were rare to begin with, and the ones I found didn’t answer. No interviews. Not a slow trickle, just silence.
Widening Instead of Narrowing
The reset came after a trip to Africa, mostly because coming back forced a clean break from however the search had been drifting before. CV rewritten, network reactivated properly instead of quietly, and a pipeline built around everything that roughly matched my field instead of the one shape I’d been holding out for.
That meant threads I wouldn’t have looked at a few months earlier. Zeit was a small, calmer team, a standard 38.5 hour week, no weekend work, but less pay and junior level responsibility for where I already was. Capgemini and MaibornWolff, the kind of corporate and consulting paths I’d worked inside of before, came back into the funnel as real options rather than fallbacks. Another startup, Suena Energy. OMR Reviews. A few more by the end that never got far enough to matter individually.
None of that was settling. It was running enough real threads at once to actually find out which one would hold up, instead of betting everything on one narrow guess landing.
What came through the far end was the exact shape I’d wanted from the start. Small, real ownership over the work. Not a compromise reached by lowering the bar, the same target reached by widening the funnel instead of narrowing it further. I’d already written about what I was actually looking for partway through this stretch, before any of it had resolved. Reading it back now, the direction held. What changed wasn’t the target, it was how many doors I was willing to knock on to reach it.
The Trial
The hiring process itself is still my favorite one to date, and worth describing on its own terms.
No technical interview. Not a shortened one, none at all. Just a conversation with one of the founders, the kind that would normally be a formality before the real evaluation starts. Here, it was most of the evaluation.
Then a two day trial. Day one was exploring the product and modeling a fictional company against it. I picked a Finnish logging company, mostly because Finnish was already living in my head from somewhere else, and spent the day understanding the platform well enough to reason about a business I’d invented for it.
Day two was building. A full feature, start to finish, from client requirements and the broader product vision rather than a spec handed to me line by line. The platform already had custom properties defined at the platform level, and the work was making those usable inside constraints across different object types, shippers, routes, products, each with their own rules about what should apply. Then presenting the build and defending the decisions behind it, out loud, to the people who’d actually have to live with them.
Nothing about that trial tested whether I could reverse a linked list under pressure. It tested whether I could take an ambiguous business problem, make real decisions inside it, and explain those decisions clearly enough for someone else to trust them. That’s a more honest test of the actual job than most interviews I sat through this year, and I’d trust it more if I were the one hiring, too.
Where It Landed
Orbit is small. Fully async, no office, remote by design rather than by accident. It leans hard on AI in how the work actually gets done, not as a slogan. Everyone I met through the process was genuinely motivated by the problem, not just employed by it.
I’ll save how it’s actually going for later. This piece is only about getting there.
The Actual Point
I knew what I was looking for early on, wrote about it while I was still inside the uncertainty, and never really changed my mind about the shape of it. Knowing that didn’t get me any closer to it.
What got me there was giving up on aiming at it directly. Widening the search until enough real threads existed to compare, letting some of them turn out to be worse fits, and trusting that the right one would still be recognizable when it showed up. It was, and the way it hired me was the first real evidence of it.
