Lead Software Platform Engineer, MLOps

Lead Software Platform Engineer, MLOpsLocal Hybrid | Cambridge, MA / San Mateo, CA | Full TimeAbout the OpportunityA category-leading Scientific Data and AI Cloud company is hiring a Lead Software Platform Engineer to own and scale AI/ML infrastructure as a multi-tenant product — not internal tooling. You'll be the technical design authority for how models, LLMs, and agents run in production, serving scientists at the world's largest pharmaceutical companies under real compliance obligations.This is a distributed systems role first, MLOps specialty second, we need a core systems engineer who has taken AI infrastructure to production at scale, not a data scientist who picked up deployment skills.What You'll DoOwn the AI/ML platform architecture: the API surface customers and internal teams build againstOwn the model and prompt lifecycle: registration, versioning, staged promotion, multi-model servingDesign the inference layer for real-time and batch workloads: routing, caching, GPU capacity planningArchitect the agentic layer above LLMs: RAG, tool/function calling, MCP-based toolingDesign multi-tenant security: PII/PHI handling, prompt-injection and tool-abuse defensesBuild evaluation infrastructure: eval harnesses, regression gates, drift detectionOwn observability and SLI/SLO/SLA definition for production AI systemsDesign for reproducibility and audit trail in a validated pharma environmentAct as SME and design authority: lead reviews, mentor senior engineersWhat You Bring10+ years building distributed, cloud-native systems in productionMulti-tenant AI/ML infrastructure shipped as an external productHands-on production LLM experience — RAG, retrieval/embedding design, prompt versioning, tool calling under real trafficExpert-level TypeScript and PythonProduction model registry/serving experience, ideally Databricks MLflowAI evaluation as a release gate — eval harnesses, regression gates, drift monitoringAWS, containerization (Docker), infrastructure-as-code (CloudFormation/CDK)Multi-tenant security design experienceNice to HaveMCP-based tooling and agent runtime experienceLLM cost/latency optimization, per-tenant cost attributionRegulated/validated environment experience: GxP, 21 CFR Part 11, SOC 2Life sciences or laboratory data background