Send flagged spreadsheet records to data engineers before training run

Your flagged spreadsheet rows sit unprocessed, causing downstream datasets to lag and models to miss recent updates. Data engineers receive processed record IDs to kick off pipelines so models stay current before the next training run.