What are Jev evals? How they work and how they compare to LLM-as-a-judge

Use Jev, TypeSafe AI's System One decision model, to judge categorical metrics in Rhesis. How it differs from LLM-as-a-judge, its limits, and setup.