Send new dataset file alerts to data engineers before runs

Your dataset files land in shared folders, causing delayed pipeline runs and stalled model retraining. The team receives prioritized alerts with file links so ingestion starts before the next scheduled run.