Manual concierge pilot
How the managed AI QA pilot works.
The pilot is a manual, fixed-scope service: choose a review type, submit the AI output, get a scoped review plan, approve the quote, then receive a structured report. No self-serve checkout.
- Intake
- Email, scoped by an operator
- Approval
- Written quote and plan first
- Output
- Structured report with evidence
How the pilot runs
Four steps with a written approval before any reviewer starts.
Choose the review type
Fact-checking, citation verification, support reply QA, RAG review, product listing QA, website QA, or an agent audit.
Submit the AI output
Paste the answer, content, listing, support reply, source documents, or agent trace, with the sources it should match.
Human review
A reviewer follows the rubric for that service and returns evidence-backed notes.
QA report
You receive a pass, revise, or high-risk recommendation, an issue log, and corrections.
Why a manual pilot
A narrow, operator-scoped service before a wide-open marketplace.
Scoping each request by hand avoids the cold-start problem of an empty marketplace, gives buyers a clearer product, and produces repeatable quality standards for each review type. The fixed services, report template, and trust policy come from that work.
No self-serve checkout is enabled from this page. Any payment-processing step happens only after you approve the written scope and quote.
Related pages
Neighbouring pages in the AI QA cluster, each with a different scope.
Send one request to start.
An operator scopes the review with you. Nothing is charged and nothing is published until you approve the plan.