Request an AI Release Evidence fit review

Leave this field empty

1 · Your team 2 · The system 3 · Release scope

Step 1 of 3

Who owns the release?

We use these details only to evaluate and respond to this B2B request.

Name *

Work email *

Company / legal business name *

Your role *

Billing country *

Product URL (optional)

I am requesting this for an organisation and am authorised to discuss its scope.

Step 2 of 3

What is already working?

Evaluation track *

General AI releaseAgent, RAG, copilot or structured workflow Research & personasBehavioural simulation or synthetic personas EducationTutor, feedback or learning workflow Not sureRecommend the smallest defensible track

System type *

AI agentMulti-step or tool-using workflow RAG / assistantGrounded search or answer workflow CopilotHuman-in-the-loop product feature Structured AIExtraction, classification or generation Behavioural simulationSynthetic personas or multi-agent system Learning systemTutor, feedback or educational content

Current stage *

ProductionPrivate betaRunnable prototypeNot runnable

Representative cases or anonymised traces *

ReadyPartialNone yet

How can we test? *

API / stagingRunner in our environmentGuided session / temporary workspaceNot sure yet

Do not enter keys here. Access is arranged only after scope and a written agreement.

Current model / orchestration stack (non-confidential)

Step 3 of 3

Which decision must the evidence support?

Target release timing *

0–30 days31–90 daysNo date yet

Offer of interest *

Metrics API€50 · prepaid 60-minute window · metrics only Evidence Baseline€2,500 · 5 business days Release Sprint€4,900 · fixes + retest Release GateFrom €1,500 after a sprint Not sureRecommend the smallest fit

The failure or uncertainty you most need to resolve *0/1200 characters

The decision you want to make at the end *0/800 characters

Important constraints (optional)0/800 characters

Behaviour, capability or construct to evaluate (optional for general release)0/1000 characters

Relevant populations, learner profiles, languages or contexts (optional)0/800 characters

Independent reference available (optional)

Anonymised observed dataExpertsTheory / published protocolExisting rubricNone yet

Memory or adaptation

NoneWithin sessionAcross sessionsPersonalisation / learningNot sure

Does the output influence a high-impact decision?

NoYes, with human reviewYes, automatedNot sure

Does the scope involve minors or another vulnerable group?

NoYesNot sure

We need an NDA before sharing any technical detail beyond this form.

Fit review requiredWe use the answers above to decide whether a bounded sprint is viable. No automated decision is made about you or your organisation.

I have not submitted credentials, source code, personal data, customer content or other restricted/confidential material, and I have the right to provide the information above. I understand that the deliverables are technical evidence and decision support—not legal advice, certification, conformity assessment or release approval. I understand that psychometric analyses describe observed system behaviour inside the agreed scope; they are not a diagnosis, formal test validation or proof of equivalence with a human population. I accept the B2B request terms and confirm I have read the privacy notice.

Submitting is free. It is a request for review—not an order, contract, payment or promise of availability.

Tell us about one release decision.

Who owns the release?

What is already working?

Which decision must the evidence support?