Design and build production-grade generative AI systems, including LLM architectures, fine-tuning pipelines, model evaluations, RAG systems, and multi-agent integrations. Optimize models for latency, throughput, token efficiency, and cost while implementing monitoring, guardrails, rollback, and enterprise security patterns. Translate client requirements into model strategies, contribute production code and design reviews, and collaborate across engineering, data, product, and client teams.