Before AI goes live, know whether it’s ready.
Answer 24 questions about one AI project, in about ten minutes. Get a clear verdict your board can act on, with the reasons and the fixes. Fixed rules, no AI in the decision.
- Organisation
- Excel Invests
- Use case
- Back-office process automation
- Jurisdiction
- US financial services
Verdict
Qualified with conditions
May go ahead within its assessed scope. 14 required actions before full qualification.
The dark line marks the pass level of 7.
A verdict, not a maturity score.
24 questions across five areas, about ten minutes. The engine places your use case on two measures, how much could go wrong and how well it is controlled, and returns one of five verdicts.
A serious governance gap blocks qualification outright, whatever the other scores say. Each area that falls short of the pass level comes with its own required actions.
Where your project lands.
- QualifiedLower exposure, strong control. The deployment may go ahead within its assessed scope.
- Qualified with conditionsAs above, with named actions to complete before full qualification.
- Production readyHigher exposure, matched by strong control.
- Ownership requiredExposure is lower, but control is weak. Name an owner who can stop the system before anything else.
- Deployment blockedHigh exposure without adequate control, or a deployment-blocking governance signal. Do not deploy.
How it works
- 1
Answer 24 questions
Five areas, plus who is accountable and how the deployment operates. No technical documentation needed.
- 2
The rules run
A fixed, versioned ruleset scores each area and checks ownership and governance conditions. No AI model is in the decision.
- 3
Receive the result
A qualification certificate and an evidence book, within minutes: the verdict, the reasons, and every answer behind it.
A record the board can sign off against.
The verdict says whether the deployment may go ahead. The required actions say what has to be true first. Both are written for the people who approve, not the people who build.
Go or no-go
One of five verdicts, with the reasons in plain English and the scope it applies to.Conditions to close
Each required action tied to the answer that triggered it, so owners know exactly what to fix before go-live.Checkable later
Fingerprints for the ruleset and the decision, so anyone can confirm the result was not changed. Why fixed rules, not an AI model.
Five areas, each scored out of 10.
An area passes without conditions at an average of 7 or above.
Data readiness
Sources, lineage and consistency across systems. The most common blocker: data quality and trust rank among the top three barriers to production AI agents for 75.9% of data leaders surveyed.1Process maturity
Which decisions are candidates for automation, and how consistently they are made today.Tooling and infrastructure
The stack, integration points and legacy constraints.People and capability
Ownership, training and leadership mandate.Governance and compliance
Policy, DPIA and named accountability.
Qualification certificate
The verdict, the reasons for it and an executive summary, with reference ID and date. A record of a rules-based result, not a certification.Evidence book
Every finding, every required action, and each of the 24 answers with its score, plus the frameworks each finding relates to.
Built for how your sector works.
Governance looks different in a school trust than in a bank. Sector assessments ask the questions that apply to you.
How it compares
| AQE | Consultant review | General AI chat | |
|---|---|---|---|
| Same answers give the same result | Yes | Depends on the reviewer | No |
| Fixed, versioned rules | Yes | Varies | No |
| Every answer and score shown | Yes, in the evidence book | Varies | No |
| Anyone can check the result was not altered | Yes, via fingerprints | No | No |
| Time to result | Minutes | Days to weeks | Instant |
| Price | £750 fixed fee, free for founding users (all prices) | Usually quoted per engagement | Free or low cost |
| Judgement on your wider context | Limited to the questions asked | Yes | Unverified |
Questions
Is this an audit or a certification?
No. AQE is a self-assessment: the result is based on the answers you give, applied to fixed rules. The certificate records that result. It does not certify your organisation or replace an independent audit.
Is this legal advice?
No. Findings point to the frameworks they relate to. They are not a finding that you are in breach of anything, and do not replace your own legal or compliance advice.
What does “Qualified with conditions” mean?
The use case sits in the qualified position, but one or more areas fall below the pass level. It may go ahead within its assessed scope, and the certificate lists the actions needed for full qualification.
Why does checkout show Clearance Diagnostic?
AQE and CLEARANCE are sister products from the same provider and share one Stripe payment account. Your payment is processed securely by Stripe and appears as Clearance Diagnostic at checkout.
Our AI is already live. Is AQE still useful?
AQE qualifies a use case before deployment. For a system already in use, CLEARANCE checks the live workflow against named obligations.
Know before you deploy.
Free for our first 20 founding users, then £750. No card required, no sales call.
- The Modern Data Company, “5 Emerging Trends in Enterprise Data & AI”, interim findings of the third Modern Data Survey, August 2026, reported by BigDATAwire. Back