Content moderation team tries Jev as a general classifier

When I started working on our content moderation system, the provider we were paying for wasn't working, and I had a hunch a general classifier could do better if we gave it a clear policy. It turned out to be a good test of how my study of Computational Linguistics and research methodology shapes how I explore a problem space with agents like @FactoryAI. When Jev was released as a general classifier, it seemed like the right one to try. Because Jev is text only, I decided we could use an off-t