• Design and implement production-grade LLM and agentic systems that ingest operational, doctrinal, and policy data and transform it into structured tasks, workflows, and resource-aware plans. • Build orchestration layers across workflows, tools, and human approvals to support a recommendation-first AI copilot with progressive, user-authorized agentic actions. • Implement RAG pipelines over doctrine, regulations, and historical data to enable constraint-aware reasoning, sequencing logic, and transparent, explainable recommendations. • Design trust, permissioning, and rollback mechanisms that adapt system autonomy based on user behavior, experience, and operational risk. • Instrument systems for auditability, traceability, and user trust, ensuring all outputs can be explained and linked back to source inputs. • Optimize AI systems for latency, cost, reliability, and explainability in real-world operational environments. • Collaborate closely with product, UX, engineering, and domain experts to translate complex operational workflows into scalable, human-centered AI systems.