The evaluation report & scorecard
What you receive at the end. It is a report and a scorecard — an independent read of the evidence. It is not a certificate, and we do not present it as one.
One view, four grades, one gate — no blended number
Each dimension is graded on its own evidence. Averaging them would let a strong capability score bury a serious safety failure, so we don't.
| Dimension | Grade | What it means |
|---|---|---|
| Safe | Fair | Behaviour under misuse, attack, and edge cases |
| Skilled | Very Good | Real capability on held-out domain tasks |
| ROI-Fit | Good | Value, reliability, and cost per successful outcome |
| Transparency | Good | How inspectable the AI's path to an answer is |
Grade scale — graded on evidence, not vibes
Release gate — the decision the grades feed
Every dimension Good or better, no open S1, and S2s within the agreed limit.
No open S1, but named issues must be fixed and re-tested before scale-up.
Any open S1, or any dimension graded Poor. Deployment is blocked until it is re-evaluated.
Two layers, never mixed: a grade per dimension describes quality, and the gate verdict describes a deployment decision derived from the exit criteria in the plan.
What the full report includes
What the grade is, and isn't
It is
An independent, evidence-backed opinion at a point in time, against a named scope, with the method shown.
It isn't
A legal certification, a guarantee, or a permanent stamp. Models drift; a grade has a date on it for a reason.