How ready is your AI program?
The method behind every engagement: twelve dimensions of readiness, each scored on cited evidence. Further down, a separate two-minute self-score from the Examiner Question Checklist. No email required.
Scored on evidence.
12 DIMENSIONS / CITED EVIDENCE
Twelve dimensions. One number your executive team can act on.
Every dimension is scored on cited evidence gathered in interviews and system reviews, and every score comes with the sequence that closes the gap. Your score is the board slide. The method is shared in the engagement, with the evidence file that supports it.
- D1 / Board & Strategy Alignment
- D2 / Data Quality & Accessibility
- D3 / Data Governance & Permissible Use
- D4 / AI Governance & Policy
- D5 / Vendor & Third-Party AI Posture
- D6 / Platform & Integration Architecture
- D7 / Member-Facing AI Readiness
- D8 / Lending & Credit-Decisioning AI
- D9 / Back-Office & Productivity AI
- D10 / Team Capability & Adoption
- D11 / Examiner Readiness & Compliance Evidence
- D12 / Measurement & Value Tracking
Two minutes. Your number.
10 QUESTIONS / 2 MINUTES
FROM THE EXAMINER QUESTION CHECKLIST
NO EMAIL REQUIRED
Separate from the Index, and yours to keep.
Ten questions drawn from our Examiner Question Checklist, built from what examiners ask about third-party risk, model risk, and fair lending. An independent self-check written by Deployed Advisory, not an official examination checklist. Answer honestly. Nobody sees this unless you choose to share it.
Do you maintain a single inventory of every AI capability in use, including AI features embedded in your vendors' platforms?
Has your board adopted an AI policy or charter, with a named accountable executive?
Does every AI vendor have a completed due-diligence file: data flows, testing evidence, contract terms?
When an existing platform switches on a new AI feature, does it pass a review before activation?
For any model influencing lending or pricing: was it validated before launch, and is it monitored since?
Could you produce specific, explainable adverse-action reasons for a model-influenced decision today?
Is the data feeding your AI documented for lineage and permissible use?
Is there an acceptable-use policy, with training, preventing staff from putting member data into public AI tools?
Does every deployed AI capability have defined metrics reviewed on a cadence, with evidence you could show?
Does your board receive regular AI reporting that shows up in the minutes?
Prefer it in writing? Email my results instead
Your score, our examiner-question checklist, and what to tackle first.
Scored below 16? A thin score is a common starting point, and much of it is documentation work.
Directional and self-reported, and separate from the Index. Answers stay in your browser unless you request the checklist.
The method behind every engagement.
Twelve dimensions we check, on evidence we review. The first engagement takes two weeks at one fixed fee, quoted before we start, and ends in a decision brief with a clear call.