Skip to content
Book a scoping call

Self-assessment · 12 questions · About 5 minutes

Twelve questions before your next AI pilot.

Pick one AI use case you're planning or already piloting. Answer each statement for that use case. You'll get a score per dimension on this page. No sign-up, and nothing is sent unless you ask for the readout by email.

For example: a procurement assistant, or invoice automation.
0 of 12

The twelve statements

0101 · Use cases and valueThis use case has a named owner and a baseline metric it should move.
0202 · Data sources and rightsWe know which systems and documents it needs, and we're allowed to use that data this way.
0303 · Platform and architectureOur data platform can carry it, and we can predict what it will cost to run.
0404 · Data engineering and operationsThe data it needs arrives on time, every day, with tests that catch failures.
0505 · Data quality and master dataEach customer, supplier, product or counterparty it touches has one ID that resolves across our systems.
0606 · Semantics and metadataThe numbers it uses have one agreed definition, and we can trace them to source.
0707 · Documents and retrievalThe documents it needs are findable, current, tagged and linked to the right records.
0808 · AI and ML engineeringWe have a test set that tells us when its answers are wrong.
0909 · Security, privacy and accessIt can only see what the requesting user is allowed to see, and sensitive data stays out of prompts and logs.
1010 · Governance, risk and complianceWe could show an auditor or regulator how it reached a decision.
1111 · People and operating modelWe have the people to build it and to run it after launch.
1212 · Roadmap and economicsIt has a budget tied to our value creation plan, including run cost.