This is so cool! We have files in Iceberg that we then move data to/from to a PG db using a custom utility. It always felt more like a workaround that didn’t fully use the capabilities of both the technologies. Can’t wait to try this out.