Create normalized fraud CSVs for pipeline consumers

Your incoming webhook payloads arrive unstructured, causing delayed fraud record ingestion and manual parsing. You get normalized pipe-delimited CSV files in S3 so downstream systems can consume records before the next billing run.