Once your traces are failing criteria, the question stops being which trace and becomes which problem. Composo clusters failing traces by similarity and names each cluster, so you triage a handful of recurring failure modes instead of scrolling a list.
What’s in the video
- The failure map - every failing trace placed by similarity, grouped into named modes
- How each mode is trending against the previous week, and which modes are new
- Opening a mode to read the traces behind it
- Promoting a mode into an evaluator, so the failure you just found is measured from now on