Partner with engineering and business teams to productionize agentic workflows and automations: turn agent prototypes into scalable, repeatable, reliable services using orchestration frameworks (e.g., LangGraph, Claude Agent SDK, Airflow, Prefect, etc) with containerized, sandboxed execution, secure tool and data integrations (e.g., MCP), and support for many concurrent users and workflows without contention. In this role, you will have the opportunity to: Own the end-to-end ML and LLM lifecycle and deployment: experiment tracking, model and prompt registries, versioning, lineage, and reproducibility (e.g., MLflow, Weights & Biases); design and operate model serving for batch and low-latency online inference, including LLM inference, with autoscaling, GPU efficiency, and performance optimization (batching, quantization, caching).