On a schedule the scenario picks up new files from Databricks and uploads them in batches to Anthropic, keeping their names.