We use a monorepo in which insights by Job (historical, graphs) is just as essential as insights by Workflow and Repo.