Evaluating chatbot/AI workflow: benchmark, verifier, LLM judge and real-world test cases

A practical note on LLM evaluation: benchmarks are just the starting point, while a real product needs test cases, verifiers, human review and metrics tied to the workflow.