What did you test your detection claim against?
A tool that claims it can spot documentation which will not survive review needs something independent to be measured against. You run a held-out record set; scoring is returned by the holder, and you receive a calibration report against credentialed human raters.
| What is read | Your tool or your team runs a held-out record set; scoring is returned by the holder |
| Turnaround | Agreed at scoping; depends on your run schedule |
| Price | $750, fixed. No hourly rate and no scope creep |
| Written for | AI assurance vendors, model evaluation teams, audit firms |
What you receive
- Licensed access to the record set for one run
- Scoring returned by the holder. The answer key is never released
- A calibration report against the human rater distribution
- A written statement of what the result does and does not support
Procurement or counsel need the fee basis, data handling and deliverable ownership on one page? That is how an engagement works.
How to start
Prefer an invoice, a purchase order, or a scoping call first? Email info@jrsstandard.com and you will get scope and an invoice in one reply.
There is no form on this page and nothing to upload here. Scope, price and turnaround are agreed in writing before any record is sent, and de-identification is agreed at the same time.
Want to see the method before you ask for anything? The Record Defensibility Check is public, ungated, and requires nothing from you.