AI Decision Review

A review of one AI system in one real business workflow, so you can make your next decision about it on evidence.

When it's useful

You have a decision coming about AI you already use: whether to expand it, renew it, improve it, restrict it or stop it. The review is built around that one decision.

How it works

1. 1

   Choose the decision

   We define the AI system, the business workflow, the leader accountable for the decision, and the choice that needs to be made.

2. 2

   Write the proof rules first

   Before reviewing evidence, we agree what would support, qualify or block the decision.

3. 3

   Record the starting point

   We record how the workflow runs today, what evidence exists, and what cannot yet be observed.

4. 4

   Review the evidence

   We examine agreed records, tests, outputs, procedures and how people use the system, and mark each material claim as supported, partly supported, unsupported, contradicted, or not observable.

5. 5

   Decide and act

   You receive the Decision Packet. Your accountable leader chooses a path: proceed, proceed with conditions, hold pending evidence, restrict or redesign, or stop. We help assign the actions that follow.

6. 6

   Re-measure after change

   When the model, vendor, workflow or people change, we recheck the evidence that change affects.

What you receive

• The decision, its owner and the proof rules, written down.
• A record of how the workflow runs today.
• An evidence ledger showing which claims are supported, partly supported, unsupported, contradicted or not observable.
• The options open to you, and the conditions attached to each.
• The blockers, who owns each fix, and what will show it is done.
• The changes or dates that should trigger a recheck.

What's in scope

One AI system, one workflow and one accountable owner. If your decision touches more systems or workflows, we scope each as its own review and agree on that before any work starts.

Built for different stages of AI use

The method stays the same at every stage. What changes is the workflow, the people involved, the evidence available and the decision itself.

Every review starts with a conversation about your decision.

Discuss your AI decision →