Dat Ngo

Dat Ngo is an AI Architect at Arize AI and the company's practitioner voice on LLM observability, evaluation, and the trace-to-improvement flywheel. His recent talks lay out how production teams should architect multi-scope eval pipelines for non-deterministic agents (span, multi-span, trajectory, and session-level evals) and walk through Arize's stated goal of automating the full observability-to-improvement cycle. Attend his session for concrete patterns on evaluating and monitoring AI agents in production.