Walk me through a time you discovered a significant error in a core dataset. How did you resolve it and prevent it from happening again?
