AI Systems, LLMOps & Agentic Workflows
AI Systems
LLMOps
RAG
Agents
Designing, evaluating, and monitoring LLM-powered systems for reliable knowledge work.
My professional profile is organized around two connected pillars: rigorous decision science and production-aware AI systems.
The common thread is trust. Whether the system estimates a treatment effect or routes work through an LLM agent, the important questions are similar: What decision is being supported? What assumptions are required? How do we evaluate quality? What failure modes matter? How do we monitor behavior over time?