The curation, harvesting and review workspace is private. The future public corpus API is a separate serving boundary.