What is cost per verified successful task?
Spend and savings
Cost per verified successful task
Cost per verified successful task = Σ attributable run costs ÷ count of verified successful runs. When the denominator is zero, or cost is missing, the value is unavailable — never 0.
One of the terms Agent Spend Optimization uses for every amount it shows: amounts in different evidence classes are never added together.
Used in Agent Spend Optimization · Methodology
Related terms.
Observed spend
Invoice-aligned, provider-reported cost for the selected UTC days.
Anomalous spend
Observed spend with features that indicate a possible problem (e.g. a day more than 2.5× its trailing 14-day median).
Potential saving
A modelled reduction if a proposed change works, under stated assumptions.
Validated opportunity
A scoped, feasible change with supporting evidence and a computable expected benefit, confirmed by a person.
Verified saving
A measured reduction in cost per verified successful task in the latest recorded experiment on production traffic, approved and verified by people other than those who proposed it and recorded the results, with no success-rate drop greater than 2 percentage points, no p95 latency increase greater than 20%, and at least 30 verified successes per window.
Cash impact
Lower usage reduces future usage-based invoices.
Last updated