Lead design and scaling of legal AI evaluation frameworks and benchmarks, build and maintain datasets, audit and score AI legal outputs, define metrics focused on reasoning and citation accuracy, and collaborate with engineering to translate legal errors into model fine-tuning feedback.