Evaluate AI-generated coding interactions, assess quality of engineering judgments, provide feedback, and define standards for AI coding agents. Comfortable making subjective judgments about engineering quality.