Evaluation Pipeline Diagram

flowchart TD
  Dataset[Golden Dataset] --> Runner[Evaluation Runner]
  System[System Under Test] --> Runner
  Runner --> Metrics[Metrics]
  Runner --> Failures[Failure Cases]
  Metrics --> Report[Regression Report]
  Failures --> Report
  Report --> Decision{Release?}
  Decision -- Pass --> Release[Release]
  Decision -- Fail --> Fix[Fix and Retest]