malahov.io

We're in

An email from Startbahn27: we're in the accelerator. Ahead — a kick-off and the first presentation, and that same evening a friend asked to join the team — and I wasn't sure at first whether to be happy about it.

By Georg Malahov

In the morning an email arrived from Startbahn27: we've been accepted into the accelerator. We submitted the application on Saturday evening — and by Monday, March 3, an answer. We need to confirm participation by email before Wednesday; on the tenth there's a ninety-minute kick-off on Zoom, and on the thirteenth — the first presentation, already offline, at a venue connected to investments — I don't even really know yet what kind of organization it is.

The email that followed: participation confirmed — Los geht's, here we go

There's now even more work — the to-do list has been joined by deadlines with concrete dates. But it's exciting: something is moving. We'll see where this all ends up.

And in the evening came the second event of the day, one I'm still digesting. A friend who does design and product vision and I talked for an hour and a half: I showed him my journal, we traded ideas, I shared my thoughts about doing a startup — and he said in plain words that he wants to take part in the project, on the team.

And here I caught myself in a strange spot: I didn't know whether to be happy. On one hand — I'm absolutely not against it. I need more than a designer — I need his brains and his connections. On the other — I could have kept the project entirely to myself and worked alone. I turned that thought over and realized that's exactly where the trap sits: alone, I can no longer keep up. The task list grows faster than I can clear it, and now the accelerator deadlines have been added on top. If a professional comes to you on his own, unprompted, and says "I want to build this with you" — it would be foolish not to take him out of greed for your own idea.

The conversation also spawned an idea that won't leave me alone. I'm currently a heavy user of Cursor — a code editor with an AI agent — and I thought: what if I set up a second project next to the code, a place to collect context — market research results, conclusions from conversations, decisions? The AI can read files and links, which means the accumulated context can be used not just for code but for running the project itself: marketing, fundraising. I'll test it on my own journal as the guinea pig. And if we're in the accelerator, everything from there gets recorded and filed into the same place.

So it turned out to be a fork-in-the-road day: in the morning the project got external confirmation that it's worth something; in the evening — its first person who came to it on his own. By Wednesday — confirm participation. By the thirteenth — figure out how to tell a room full of strangers, in a few minutes, what I've been dictating into this journal for months.