SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Eli Lilly's Global DIA (Digital Innovation & Analytics) organization is seeking an AI Solution Architect to own end-to-end AI and analytics architecture across the Manufacturing & Quality (MQ) technology portfolio. This is a strategic, hands-on leadership role for a senior technologist with deep expertise in pharmaceutical manufacturing, enterprise AI platforms, and GxP-regulated system design.
Key Responsibilities:
AI Architecture & Technical Leadership: Define and own AI/analytics solution architecture across the MQ DIA portfolio, covering data ingestion, RAG pipelines, agentic orchestration, MCP server ecosystems, model serving, and GxP-compliant deployment. Lead Architecture Review Boards (ARBs), produce Architecture Decision Records (ADRs), and enforce design standards across squads and platforms. Evaluate and select AI/ML frameworks, cloud services, and data platforms aligned to Lilly's enterprise direction. Translate executive strategy into multi-quarter technical roadmaps with phased deliverables, resource estimates, and risk registers.
Agentic AI, RAG & Advanced Analytics: Architect multi-agent orchestration systems (LangGraph, AutoGen, CrewAI) for manufacturing workflows including deviation investigation, complaint triage, CAPA recommendation, and batch release support. Design enterprise RAG architectures across QMS, MES, and regulatory knowledge bases with retrieval fidelity, hybrid search, and audit-ready provenance. Define MCP server architecture standards for exposing manufacturing APIs (Veeva, MES, LIMS) as structured LLM-consumable tool interfaces. Architect real-time analytics on OSI PI / Kafka streams for predictive quality, OEE optimization, and process excursion detection.
Cloud & Data Platform: Own cloud architecture strategy across AWS, Azure, and Databricks, ensuring security posture, GxP compliance, and cost governance. Design the data lake integration layer using medallion architecture (Bronze/Silver/Gold), data contracts, schema registries, and lineage governance on Delta Lake. Define MLOps architecture including model registry (MLflow), drift detection, retraining pipelines, and model lifecycle governance. Govern MuleSoft + MCP integration patterns with clean separation between legacy integration investments and new AI tooling layers.
GxP Compliance & AI Governance: Serve as AI governance authority for model explainability, audit trails, human-in-the-loop gates, and regulatory defensibility of AI-assisted decisions. Lead CSV framework design for AI/ML systems under 21 CFR Part 11, EU Annex 11, and GAMP 5, including IQ/OQ/PQ templates and risk-based validation strategies. Engage Global QA, Regulatory Affairs, and IT Risk to qualify new AI capabilities for GxP manufacturing environments.
Team Development & Stakeholder Leadership: Mentor distributed AI engineering and platform architecture teams; define engineering standards and review processes. Represent MQ DIA in enterprise AI governance forums, architecture councils, and with executive stakeholders across Manufacturing, Quality, and Regulatory. Partner with HR to define hiring profiles and skill frameworks for the MQ AI engineering community.
Requirements:
Education: Bachelor's in Computer Science, Engineering, Data Science, or related field; Master's or PhD preferred.
Experience: 10–15 years in AI/ML systems, solution architecture, and enterprise data platforms; minimum 5 years as principal/lead architect. 3 years in pharmaceutical, biotech, or regulated manufacturing with GxP / 21 CFR Part 11 exposure.
Technical Skills: Expert command of RAG, agentic orchestration, MCP, embeddings, and prompt engineering in production environments. Advanced proficiency in AWS (SageMaker, Bedrock, EKS, Redshift), Azure (Azure OpenAI, Azure ML, Synapse), and Databricks (Delta Lake, MLflow, Mosaic AI). Strong knowledge of ISA-95/88 manufacturing standards, pharmaceutical QMS concepts (CAPA, OOS, batch disposition), and pharmacovigilance data flows.
Preferred: Hands-on experience with Veeva QMS, QBDVision, OSI PI, MES (Tulip, Werum PAS-X), or LabVantage / STARLIMS. Familiarity with AI Gateway platform, Lakehouse/Databricks environment, or MuleSoft integration architecture. Knowledge of FDA AI/ML-based SaMD guidance, ICH Q8–Q10, and data mesh / ontology engineering. Certifications: AWS Solutions Architect Pro, Azure Solutions Architect Expert, Databricks ML Professional.
