30-day Private AI Proof
One team. One document set. One measurable problem.
Use a controlled proof to determine whether an IntegAI Policy Advisor can help with one defined organisational job before considering wider deployment or licensing.
Produce a decision, not a demonstration theatre.
The proof begins with an agreed use case and ends with documented evidence against named acceptance criteria.
The closing decision is to proceed, revise or stop.
Included
- ■Use-case definition and named owner
- ■Approved source register
- ■Configured role and operating boundaries
- ■Agreed test scenarios
- ■Qualification, refusal and escalation cases
- ■Evidence requirements
- ■Defined support window
- ■Review of results against acceptance criteria
- ■End-of-proof decision record
Boundaries
- ✕One named team and use case
- ✕Approved data only
- ✕No unrestricted public access
- ✕No production dependency
- ✕No automated legal, regulatory or employment decisions
- ✕No compliance certificate or regulatory approval
- ✕No production managed-hosting commitment
- ✕Scope, access, support and deletion or return terms agreed before commencement
A good first proof is narrow enough to control and useful enough to measure.
It has repeatable questions, an approved document set, a named accountable person and a clear boundary between guidance and decision authority.
What happens next
Qualification review
Use-case and data boundary
Proof agreement and acceptance criteria
Controlled configuration and access
Agreed scenarios and evidence capture
Proceed, revise or stop decision
Buyer-owned, embedded, licence or strategic discussion if justified
Describe the controlled use case.
Tell us about the team, document set, problem, timeline and commercial route. Do not include confidential, personal or commercially sensitive information.