Engineer - MLOps & Scientific Platforms - Data Foundry

Position Summary

We are seeking an Engineer - MLOps & Scientific Platforms - Data Foundry to operationalize Data Foundry’s scientific tools and analytical methods into actionable prototypes, building ML deployment pipelines, model serving infrastructure, API layers, and observability guardrails.

• Build and maintain end-to-end ML deployment pipelines (experiment tracking, model versioning via MLflow/W&B, containerized serving, automated retraining triggers).
• Develop model registry and feature engineering pipelines for computational scientist access.
• Implement monitoring/alerting for data pipelines, APIs, ML models, and agentic systems; ensure reliability and performance.
• Build dashboards/metrics (pipeline execution, API latency, token usage, prediction quality, system health).
• Establish structured logging and tracing for debugging and performance optimization.
• Deploy predictive/analytical methods with versioning, structured error handling, and response-time guarantees; product ionize via partnerships.
• Build serving for synchronous and asynchronous/batch agent workloads.
• Define API contracts, documentation standards, and testing frameworks for robust, consumable scientific tools.
• Build/operate cloud-native serving infrastructure (AWS/Azure/GCP), using containers, Kubernetes, and IaC.
• Develop CI/CD for ML models (validation, A/B testing, canary, rollback).
• Integrate model serving with versioned, properly formatted training/inference data.
• Partner to expose tools via REST APIs and MCP-compatible endpoints; collaborate on latency/throughput and graceful degradation.

• BS or MS in CS, Data Science, ML, Bioinformatics, Computational Biology, or related field.
• 3+ years in MLOps, ML engineering, or scientific platform development.
• Authorized to work in the US full-time;
  Lilly does not sponsor visas.

• Pharmaceutical/biotech research experience.
• Strong Python; ML frameworks (PyTorch/Tensor Flow/scikit-learn) and lifecycle tools (MLflow/W&B/Kubeflow or similar).
• Production model serving experience (containers, REST/gRPC, operational monitoring).
• Cloud (AWS/Azure/GCP), Kubernetes, CI/CD automation.
• Experience operationalizing scientific/computational models (e.g., cheminformatics, bioinformatics, QSAR, PK/PD).
• Model monitoring/drift detection/retraining systems.
• API gateway/event-driven/service mesh familiarity.
• Feature stores/data versioning (DVC) and large-scale experiment tracking.
• AI agent frameworks (MCP/Lang Chain) and/or programmatic AI-invoked APIs.
• C/C++/CUDA/GPU-accelerated computing and containerizing HPC workloads (Singularity/Apptainer).