I’m currently optimizing our automated content pipeline. One issue we’ve faced is managing the massive amount of metadata for different profiles during testing, which leads to flaky errors when multiple parallel workers mutate a single shared staging database. To eliminate this friction, we...