Send new dataset file alerts to data engineers before daily runs

Your raw dataset files land in shared folders without context, causing delayed model training and missed ingestion windows. Data engineers receive concise file alerts so pipelines can start before the next scheduled run.