AI Teams

👥 AI Teams 101

What is an AI Team?

An AI team is a set of AI roles working together on one outcome. Instead of asking one general assistant to do everything, you give different roles different jobs: researcher, critic, strategist, editor, risk reviewer, planner or operator.

Browse AI Teams AI Workflow Guide Request a Workflow

Why AI teams matter

One AI can draft, analyse and review, but it often blends those jobs together. AI teams make the roles explicit, which improves challenge, structure and quality.

One general assistant

Everything is handled by one voice in one pass.

Research this, write the plan, check the risks, improve the tone and give me the final answer.
  • It may agree with itself too easily.
  • Critique and creation get mixed together.
  • Risk review can be shallow.
  • The output can sound polished before it is actually strong.

AI team

Each role has a clear job and the output moves through a simple process.

Researcher gathers options → Strategist creates plan → Critic challenges it → Risk Reviewer checks exposure → Editor makes it clear
  • Better challenge and second-order thinking.
  • Clear separation between drafting and review.
  • Reusable role structure.
  • Stronger outputs for complex work.

Common AI team roles

An AI team is usually built from specialist roles. You do not need all of them. Pick the roles that match the problem.

🔎

Researcher

Finds information, summarises source material, extracts facts and identifies what is missing.

🧭

Strategist

Turns information into options, direction, trade-offs and a recommended path.

🧨

Critic

Challenges weak logic, blind spots, assumptions and overconfident conclusions.

🛡️

Risk Reviewer

Checks legal, security, financial, operational, reputational or safety risks.

✍️

Editor

Makes the output clearer, sharper, shorter and better matched to the audience.

📦

Operator

Turns the plan into steps, checklists, tickets, handoffs or automation-ready instructions.

✅

Quality Gate

Checks the final output against the brief, pass criteria and human approval points.

👤

Human Owner

The person who approves, rejects or takes responsibility for the final decision.

How an AI team works

The best AI teams are simple. Define the roles, run them in order, then make the handoff clear.

1. Load context Give the team the background, goal, audience and rules.
2. Assign roles Pick the roles needed for the outcome.
3. Generate One or more roles create the first version.
4. Challenge Critic, risk or quality roles test the work.
5. Finalise Editor or operator turns it into usable output with approval points.

A simple AI team example

This is what an AI team might look like for making a decision from messy information.

Team setup

🔎
ResearcherSummarises the source material and lists known facts, unknowns and assumptions.
🧭
StrategistCreates options, trade-offs and a recommended path.
🧨
CriticChallenges the recommendation and looks for weak logic.
🛡️
Risk ReviewerFlags operational, security, legal or reputational exposure.
✍️
EditorTurns the result into a clear decision brief for humans.

Team prompt example

AI Team: Decision Review Team

Goal:
Turn the supplied notes into a decision-ready brief.

Roles:
1. Researcher: extract facts, unknowns and assumptions.
2. Strategist: propose 2-3 options with trade-offs.
3. Critic: challenge the options and identify weak logic.
4. Risk Reviewer: flag operational, security, financial and reputational risk.
5. Editor: create the final decision brief.

Output:
- Situation
- Known facts
- Options
- Recommendation
- Risks
- Assumptions
- Decision needed
- Human approval required: Yes / No

Rules:
Do not invent missing facts.
Mark uncertainty clearly.
Separate evidence from opinion.

When should you use one?

Use an AI team when one role is not enough. They are best for complex work that benefits from multiple perspectives, critique and review.

Use an AI team when…

  • The task has research, planning, review and editing stages.
  • You need challenge, critique or risk review.
  • The output will influence a decision.
  • Different perspectives improve the result.
  • You want a repeatable team process for complex work.

Use something simpler when…

  • A normal prompt gets the job done.
  • The task is quick and low-risk.
  • The roles would just repeat each other.
  • You are adding process theatre instead of value.
  • The team output takes longer to read than the actual work.

AI team types

Different teams suit different jobs. Build around the outcome, not the fanciest job titles.

📊

Decision Team

Researches options, challenges assumptions, reviews risk and produces an approval-ready brief.

Browse decision teams
📝

Content Team

Plans, drafts, critiques, edits and quality-checks content before publishing.

Browse content teams
🔐

Risk Team

Reviews proposals, code, processes or plans for security, governance and operational exposure.

Browse risk teams

Where to go next

AI teams work best when paired with context files, skills, workflow packs and quality gates.

👥

AI Teams

Browse reusable teams for decision support, writing, research, risk, operations and automation.

Open AI teams
🧭

AI Workflow Guide

Understand how teams fit with prompts, skills, context files, packs and gates.

Read guide
📦

Workflow Packs

Use AI teams inside larger workflow packs for repeatable end-to-end outcomes.

Browse packs

Need an AI team for your workflow?

Describe the outcome you want and get routed toward the right AI team, skill, context file, quality gate or workflow pack.

Request a Workflow