How would you structure an ETL pipeline to ingest raw, unstructured data and load it into a structured cloud data warehouse?
