Anyone who has been on call for a Databricks job knows the pattern. A pipeline fails, someone pulls the run logs, screenshots a DAG from the Jobs UI, pastes a stack trace into a doc, and calls it a postmortem. The information is technically all there...
