Your question is Handle Incomplete Pipeline Data. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You're working on a data pipeline where source data can arrive with missing fields, duplicate records, or conflicting values across systems. You want a practical approach for keeping downstream reporting usable while preserving traceability back to the raw inputs.
How would you handle a dataset that is incomplete or inconsistent?