Parks engineers receive weekend CSV drops of reservation exports and need them in a Delta bronze table. What should they use?

Parks engineers receiving weekend CSV reservation exports need a straightforward file-to-Delta path: a notebook or jobified notebook that reads those drops into a Delta bronze table. For static weekend file landings that already arrived as CSV, notebook or batch file reads are appropriate and keep bronze structured for later refine into occupancy and revenue metrics. That fits civic file-drop patterns without forcing a streaming bus onto files already sitting in a landing folder.

A pipeline loading the drops into a gold aggregate table skips bronze entirely, so there is no faithful copy of the original rows to replay from if the aggregation logic ever changes—the stem asks for a bronze landing, not a jump to curated aggregates. A notebook transforming the drops into a silver table has the same gap in reverse: it does cleansing work that belongs downstream of bronze, but there is still no raw bronze table underneath. An Event Hubs stream configured for the weekend CSV batch load misapplies streaming infrastructure to files already landed as static drops—Event Hubs is built for continuous streams, and wrapping one around a folder of files adds machinery this pattern never needed.

Exam caveat: choose ingest technology by arrival pattern—weekend CSV drops are batch or file ingest; Event Hubs is for event streams, not a mandatory wrapper around files already in a folder. Operational check: place sample weekend reservation CSVs in the landing path, run a notebook that reads them into a Delta bronze table with an explicit schema, validate row counts and reservation keys, handle malformed rows intentionally, and schedule the notebook or job for Monday morning after weekend drops.