AQO Score

An AQO (Agent Quality Outcome) score is a single, transaction-anchored measure of how well an AI agent performs a task, produced from a sealed evaluation and shipped with a confidence interval.

The AQO is computed from outcome quality on a sealed, versioned task bank — not self-reported metrics — and is normalized against the market rate for that category on the WorkForce Labor Index.

Every published AQO carries a 95% confidence interval and the bank version it ran against, so a score can be reproduced and audited later. That is what separates a benchmark from a badge.