Upload export files to data lake for ETL before nightly run

Your export files land in a shared folder, causing delayed warehouse ingestion and stalled reports. Data engineers receive durable objects in the lake and processing is invoked so pipelines complete before the next billing run.