What an AI champion actually does (and the two traits that matter more than coding).

The most common reason teams hesitate to build AI in-house is a quiet one: “we are not technical enough.” They picture hiring an engineer, or worse, becoming one. Neither is true. What you need is a champion, and the role looks nothing like a developer job.

Here is what an AI champion actually does, the two traits that matter more than any technical skill, and how to spot the person on your team who fits.

The role, not the resume

A champion is the person who leads AI inside your business. They try things, build small apps for real problems, and show the rest of the team what is now possible. They do not write code from scratch. The modern tools do the heavy lifting; the champion points them at the right problems.

The two traits that matter more than coding

• Curiosity.The willingness to poke at a tool, break it, and try again. A good champion enjoys the question “I wonder if I could automate this.”
• Closeness to the messy work. They know where the time actually goes, because they live in the day-to-day. That instinct for which process is worth fixing beats technical depth every time.

Notice that neither of those is on a developer CV. The best champions are often operators, not engineers.

A normal Tuesday, three months in

By month three, a champion's week looks like this:

• They spot a repetitive task eating an afternoon and decide it should be an app.
• They build a rough but working first version that same week.
• They put it in front of the person who does the job and fix what is clunky.
• They move on to the next obvious thing.

No grand roadmap. Just steady, useful building, starting with the most annoying problems first.

The real time cost

A champion does not quit their day job to do this. It is a few hours a week, named honestly. The early weeks are mostly learning; after that it is building in the gaps. The payoff is that the work they automate hands time back to the whole team, including themselves.

Why we train two people, not one

Every Setup trains a champion and a partner. The second seat is not an upsell, it is insurance. If your champion gets busy, takes leave, or moves on, the capability does not walk out with them. Two people also learn faster, because they have someone to think out loud with.

How to spot yours

You are probably already picturing someone. It is the person who automates their own spreadsheets, who asks “why do we still do this by hand,” who tried AI on a work problem without being told to. The job title does not matter. If you do not have an obvious candidate, that is fine: helping you find one is part of week one.

If a curious operator with a few spare hours can hold this role, then building AI in-house is within reach for most teams, and a six-figure engineering hire is the slower, costlier path. And if you genuinely have nobody to spare right now, that is exactly what Run is for: we hold the workflow until you do.