Nine independent reviewers score your document without seeing each other's work. One reconciler resolves the disagreement into a verdict you can act on.
Each of the nine seats reviews the proposal in isolation. No seat sees another seat's notes or score before submitting its own.
The reconciler only resolves disagreement across the nine scores into the two composites. It does not introduce new judgments of its own.
Your document goes in. A two score verdict with ten dimensions comes out.
| Axis | Single model review | VerdictTank |
|---|---|---|
| Reviewer count | One model, one pass | ✓ Nine independent reviewers |
| Scoring | One flattened score | ✓ Two scores, ten dimensions |
| Disagreement | Never surfaced | ✓ Reconciled by a dedicated seat |
| Blind spots | Inherits the model's own | ✓ Diluted across nine independent reads |
| Revision loop | One shot, no structure | ✓ Submit, score, revise, resubmit |
Every submission, regardless of industry or stage, returns the same shape: two scores, ten dimensions, one verdict.
Upload your proposal draft in its current state, gaps and all.
Nine reviewers score independently, then the reconciler issues a verdict.
Two scores, ten dimensions, and the specific gaps holding you back.
Close the flagged gaps and run it again to watch the scores move.
Your documents are processed by 9 independent review providers. 8 of the 9 do not train on your content. See the data-handling disclosure for the one exception and our retention policy.
We say this plainly because a review product that hides how it handles your document has already failed the trust test before the first score comes back.
Submit a draft and get a verdict back with the two scores and ten dimensions behind it.