01Responsibilities
Own end-to-end ML architecture, feature engineering, and pipeline design decisions for key commercial analytics initiativesDesign and review ML architectural decisions with stakeholders, setting patterns that other engineers build onOwn CI/CD pipeline orchestration, deployment (Docker/Kubernetes/Prefect), and production monitoring across multiple projectsApply software engineering rigor and best practices to ML systems, including CI/CD, automation, and testingOptimize model hyperparameters and evaluate model performance, robustness, and explainability across production ML systemsDesign and build production agentic AI systems using frameworks such as LangGraph multi-step reasoning, tool use, and orchestration across complex workflowsOwn the LLMOps practice for initiatives you lead: prompt versioning, evaluation pipelines, cost/latency monitoring, and guardrails for production LLM applications (Claude or similar)Architect retrieval-augmented generation systems and integrate vector databases (e.g., Pinecone) for semantic search and retrieval at production scaleServe as the primary technical point of contact with Lilly's Agentic AI engineering team, defining technical contracts, APIs, and shared SLAs as programs adopt agentic capabilitiesSet and document human-in-the-loop boundaries in partnership with Data Science and business stakeholdersProvide informal technical oversight for 2-3 more junior engineers reviewing designs and code, and unblocking hard technical problems, without formal people-management responsibilityCoordinate with diverse stakeholders such as Data Scientists, software engineers, and infrastructure teams to design the most optimal ML and agentic pipelinesRequired
13+ years of demonstrated expertise building ML/AI systems in production including model versioning, data/model lineage, monitoring, deployment, optimization, scalability, and automated pipelines with substantial recent depth in generative AI and agentic system development, not just brief exposureKnowledge of architectural design and implementation of end-to-end ML and agentic AI solutionsStrong knowledge of core ML frameworks (scikit-learn, PyTorch, TensorFlow, Keras, or equivalent) and the ability to understand and extend the modeling work of Data Scientists into production-grade systemsStrong knowledge of Python and PySpark for large-scale data processing; working knowledge of .
