Why AI doesn’t start with AI.

Most AI projects start with the technology: a tool, a licence, a vendor pilot. And that’s exactly where they get stuck.

The pattern is familiar. An organisation buys an AI platform, rolls it out, and six months later almost no one uses it. Not because the technology doesn’t work, but because the process underneath was never clear. You can’t rescue a messy process by putting a smart model on top of it. Then you’re just automating the mess.

AI doesn’t start with AI. AI starts with process design. First understand how the work really flows: which steps, which decisions, which exceptions. Only when that’s clear do you know where AI adds value — and, just as important, where the human must keep deciding.

First understand the process. Only then automate.

That sounds slow, but it’s actually faster. Design the process first and you build something that keeps working and can be extended. Start with the tool and you build something no one trusts — and you get to start over in a year.

That’s why we reverse the order. We read the existing process, reduce it to clear components, and decide together where AI takes over the repetitive steps. The technology is the final piece, not the starting point.

Arjan van Unen — CMO, Model My Context