What you actually do

You solve problems by meeting deadlines, not by filling in the backlog. Industrial companies describe processes to you that have been running this way for years – in service, sales, operations. Your work begins the moment someone says, "That won't work for us." You analyze this in meetings with the people affected.

You build the first version yourself. A mock-up, a demo, a small MVP – fast enough for the client to make a decision. Not production-ready, but convincing. Your tool for this is AI coding agents, and you're faster at it than most developers you know.

Others build the depth. Once the decision is made, the production-ready implementation goes to our AI agents and our tech teams, which you manage. You maintain the line between what was promised in the deadline and what the final result is.

You also decide what doesn't get built. A technically sound use case that doesn't offer any economic benefits won't be implemented – even if the customer wants it. You have to be able to justify this decision in meetings with the people who are paying for it.

That's you

You've already built something that wouldn't exist without you. A product, a company, a side project, an organization. We're not interested in how big it was—we're interested in the fact that it exists and that someone other than you uses it.

If something is unclear to you, you build a version of it. You don't write a concept about an idea, you make it visible.

You'll get faster in conversation, not slower. If someone throws you an unfinished problem, you'll find the structure in talking it through. And you can handle it if someone in the room contradicts your initial idea.

You ask about the business before you ask about the architecture. How often does this happen? What does it cost today? Who decides this? Only then does it become technical.