Use cases
Common workflows for reproducing, comparing, and verifying AI executions.
Reproduce a failing run
An AI execution failed in a way that is difficult to reproduce. TraceFlowLens records the execution context, including inputs, outputs, timing, tool activity, and errors, so the same run can be replayed and investigated.
Regression-check a prompt or code change
A prompt, tool, model call, or application change may alter execution behavior. Replay a previous run against the updated application and identify the exact step where the behavior changed.
Verify a fix
Compare the original failing execution with a new run after the change. TraceFlowLens highlights differences in execution structure, outputs, and status, helping confirm that the intended issue was resolved without introducing unexpected changes.