An AI-enabled team needs more than people with ideas.
Ideas are cheap now. The team needs the mix of people required to carry an idea through the work.
It needs builders who can make something real. Not just talk about a future state, but create the artifact, workflow, prototype, tool, test, or system that lets the team learn.
It needs operators who understand where the work actually breaks. Operators know the exceptions, delays, workarounds, approval paths, and human behaviors that do not show up in the clean process map.
It needs designers who can make new work usable. If people cannot understand, trust, or repeat the workflow, the capability does not land.
It needs storytellers who can explain why the change matters. Not marketing gloss. Clear narrative. What problem is being solved? What changes for the person using it? Why now?
It needs domain experts who know what cannot be wrong. Every domain has facts, risks, norms, and exceptions that a general model will not respect by default.
It needs technical architects who can see what will fail later. The demo can work and the system can still be the wrong shape.
It needs educators or change agents who can turn one person's experiment into a pattern other people can adopt.
It needs customer-facing translators who can explain the work in the language of the people affected by it.
On a small team, one person may cover several of these modes. That is normal. The point is not to create a large org chart. The point is to check whether the team has the capabilities needed to move from spark to shipped value.
If everyone is an idea person, nothing ships.
If everyone is a builder, the wrong thing may get built quickly.
If everyone is a strategist, the work may never survive contact with a real workflow.
If everyone is a specialist, the handoffs may become too brittle.
The best teams mix depth and range. They have people who can own a craft and people who can see across the work. They know when to go deep and when to translate.
The question is not "do we have an AI team?"
The question is whether the team can make useful AI-enabled work real, trusted, adopted, and maintained.