---
title: AI-Native Teams: Four Capabilities, Not Four Roles
url: https://agile-agilist.com/an-ai-native-team-isnt-built-from-four-roles-its-built-from-four-capabilities-and-the-difference-is-why-the-team-owns-the-outcome-together/
description: AI-Native Teams organize around four capabilities — Product, Builder, Domain Expert and AI — not four job titles. People move between them, which is why the team owns the outcome together.
---

# AI-Native Teams: Four Capabilities, Not Four Roles

Every AI-Native Team brings four capabilities to the work: Product, Builder, Domain Expert and AI.

The word that matters is *capabilities* . These are not four fixed roles, and that distinction changes how the team works.

### What an AI-Native Team is

The framework describes AI-Native Teams as teams that integrate human creativity with AI to deliver customer and business value. They are typically smaller than traditional teams — three to seven people, augmented with AI — and they work in a flow-based rhythm of three activities: Align, Sense and Respond.

### The four capabilities

- **Product**  defines why the work matters and keeps the team connected to customer needs.
- **Builder**  turns intent into working systems using AI. This is where most team members spend most of their time.
- **Domain Expert**  brings the specialized customer, operational or business context that makes a general AI proposal fit your situation.
- **AI**  provides automation and continuous analysis, working under human supervision and inside clear guidelines.

### Why capabilities, not roles

People move between these capabilities. You might bring Builder capability to one task and Domain Expert capability to the next. A product-minded engineer brings Product to a discovery session and Builder to the implementation that follows.

A role model asks, “whose job is this?” A capability model asks, “what does this task need, and who on the team can bring it?” In a team of three to seven, with AI taking on more of the execution, the second question is the only one that scales.

That is why the team owns the outcome together rather than each person owning a role. Nobody can say “my part is done” while the outcome hasn’t moved.

### What a capability model asks of you

**Check coverage, not headcount.**  A team can have seven people and still lack Domain Expert capability. It can have four people and cover all four well.

**Treat AI as a capability with rules.**  AI is named as one of the four, but it always works under human supervision. Team members review AI output for quality and bias.

**Measure the team on outcomes.**  If people are rewarded for their role’s output, they will optimize their role. Shared outcomes keep everyone pulling the same way.

### Three questions for your teams

**1. Which of the four capabilities is weakest on each of our teams?**

**2. Can our people move between capabilities, or do job titles stop them?**

**3. Is each team measured on a shared outcome, or on what each role produced?**

The next four pieces take each capability in turn: Product, Builder, Domain Expert and AI.

The world has changed. The leaders who notice will be the ones the next decade is built around. I write on Thursdays at agile-agilist.com.

#### Sources

- Scaled Agile, Inc. — AI-Native SAFe framework guidance: AI-Native Teams
- Scaled Agile, Inc. — AI-Native SAFe Session 3: AI-Native Teams, Roles, and ARTs