Copy updated dataset files for data engineers before pipeline runs

Your updated dataset files land in cloud Drive and risk stalling model training and ETL jobs. Data engineers get durable copies in a central data lake so pipelines run reliably before the next training cycle.