Product / Evaluations

Evaluate the decisions agents make.

Agents make assumptions when context is missing. Those decisions are checked against policy, and patterns of incorrect behavior surface across runs.

Policy conditions change. Were past decisions still valid when they were made?

A threshold is exceeded. A deadline passes. A boundary moves. Each decision is checked against the conditions active at the time.

Geographic Boundary
EU
Within approved region
Volume Threshold
800K records
Within 1M limit
Retention Deadline
45 days old
Within 90-day window
approved
All conditions met → Operation approved

Evaluations

Assumptions

Values checked against policy

An agent fills a gap with an assumed value and moves on. That assumption is checked against the policy that should have applied, and repeat offenders surface across runs.

Drift

Caught before it compounds

Agent decisions diverge from what policy intended. Decisions are tracked across runs, so drift shows up as a trend while it's still small.

Output review

Policies as evals

New outputs need checking before they're trusted. The same policies that guide decisions run as checks against output, flagging anything outside policy before downstream use.