What this tool can actually measure from its own data - not a first-pass approval rate or a dollar rework figure. Those need real jurisdiction approval/rejection outcomes this app doesn't have visibility into; reporting them from flag counts alone would be a guess dressed up as a metric. Precision below only reflects flags reviewers have marked via the status control on the flags page - it stays blank until reviewers start using it.
| Project | Revisions | Cycle time | Flags caught | High-severity, still open | Reviewed | Reviewer-confirmed precision |
|---|---|---|---|---|---|---|
| Demo: Chicago Labeled Conflict Set | 1 | - | 2 | 0 | 2 | 100% (2 real / 0 false positive) |
| Demo: Cross-Discipline Clash Set | 1 | - | 8 | 3 | 0 | not yet reviewed |
| Demo: Data Center Campus (Full Set) | 1 | - | 47 | 11 | 0 | not yet reviewed |
| Demo: San Diego Labeled Conflict Set | 1 | - | 2 | 0 | 0 | not yet reviewed |
| Riverbend Family Dental Clinic | 3 | 0 day(s) | 48 | 0 | 0 | not yet reviewed |
| Test: Chicago Dentist | 1 | - | 39 | 0 | 0 | not yet reviewed |