• Design and implement comprehensive evaluation frameworks that reflect real-world task success for agentic systems, with a focus on human+AI collaboration outcomes • Build benchmarking pipelines that capture nuanced success indicators including trust calibration, intervention frequency, and agent handoff quality • Lead development of observability tools and instrumentation for analyzing agent behavior in production • Translate complex qualitative and quantitative signals into actionable insights that inform model iteration and product prioritization • Collaborate with researchers, engineers, and product teams to align evaluation methodologies with business and user goals • Own benchmarking infrastructure that enables reproducible, scalable evaluation across AI initiatives • Champion rigorous experimental design and statistical analysis across teams to ensure consistent and meaningful measurement standards