You're troubleshooting a data pipeline issue and suspect it may be tied to a recent product change. You want to separate a true pipeline regression from a source-side behavior change before making fixes.
What would you do if you suspected the issue was caused by a recent product change?
Timing correlation between anomaly onset and release timestampChanges in Airbyte raw payload shape, nullability, or volumeConnector logs, sync metadata, and downstream dbt test failuresDifferences between source behavior change and pipeline logic regression