Agentic AI Data Engineer - CMC Data Integration

At Lilly, the work is demanding because patients are waiting. We unite caring with discovery to help make life better for people around the world, knowing that every decision, every detail, and every day matters.

Headquartered in Indianapolis, Indiana, our over 50,000 employees around the globe take on complex challenges to discover and deliver life-changing medicines, strengthen how health is understood and managed, and support the communities we serve. This is hard, urgent, selfless work—but it’s work worth doing.

If you’re driven by purpose and ready to bring your best to work that truly matters for patients, we invite you to join us. Overview: The Bioproduct Research and Development organization strives to deliver creative medicines to patients by developing and commercializing insulins, monoclonal antibodies, novel therapeutic proteins, peptides, oligonucleotide therapies, and gene therapy systems.

This multidisciplinary group works collaboratively with our discovery and manufacturing colleagues. We are seeking an AI Data Engineer to build the data ingestion infrastructure and a unified data model that underpins the modernized CMC Data Backbone.

This is a hands-on engineering role with design influence — you will write production-quality pipelines, define CMC data schemas, and work directly with scientists and digital architects to ensure data from internal LIMS/ELN systems and external CDMO partners flows reliably into a single data backbone. You will work with a team of engineers and data scientists.

You will have the autonomy to own your components end-to-end. If you want hands-on experience at the intersection of pharmaceutical science and modern agentic AI data engineering — agentic pipelines, document AI, GxP-compliant data infrastructure — this is the role to build that foundation.

Key Responsibilities: Agentic Pipeline Components: Implement individual agent components (e.g., document extraction agent, schema mapping agent, validation agent) within the established orchestration framework (LangGraph, LlamaIndex, or equivalent) Write tool-calling logic, handle failure modes, and ensure each agent component is testable and observable with instrumented logging of inputs, outputs, and intermediate decisions Iterate on agent behavior based on real data performance; work with the senior engineer to identify and resolve failure patterns Participate in validation and qualification activities for AI-assisted workflows, supporting documentation that demonstrates computational tools reflect scientific intent Human-in-the-Loop (HITL) Workflow Implementation: Build review queues and flagging logic that surface low-confidence or out-of-specification extractions to scientific reviewers for approval before data is loaded Implement routing logic that captures reviewer decisions, logs outcomes with full audit trail, and reintegrates approved data into the pipeline per 21 CFR Part 11 electronic records requirements Tune flagging thresholds based on feedback from scientific owners; maintain and improve HITL logic as new data sources are onboarded Data Ingestion & Pipeline Engineering: Design and build AI-assisted ingestion pipelines that extract and structure the data from unstructured CDMO/CRO data sources: PDFs (Certificates of Analysis, batch records), Excel files, and vendor portal exports Implement validation, reconciliation, and exception-handling logic to ensure data completeness and integrity before loading Build monitoring and alerting for pipeline health, data quality, and ingestion failures Design a data quality framework with automated checks, rejection handling, and audit trail logging. Develop reusable pipeline templates and schema documentation that reduce onboarding time for new CDMO partners Required Qualifications: MS in Computer Science, Computer Engineering, Data Engineering, or related technical field with 1–2 years of relevant experience; OR BS in Computer Science or Computer Engineering with 3–5 years of hands-on data engineering experience.

Proficiency in Python and SQL; ability to write, review, and own production-quality code. Demonstrated experience building ETL/ELT pipelines from unstructured or semi-structured sources (PDFs, Excel, JSON, XML).

Hands-on experience building LLM-powered applications: retrieval-augmented generation, tool-calling, multi-step orchestration, or equivalent agentic patterns. Hands-on experience with cloud data platforms: Azure (Data Factory, Databricks, Fabric) or AWS (S3, Glue, Lambda, Redshift).

Solid understanding of relational data modeling, schema design, and data normalization principles. Familiarity with data orchestration tools (Airflow, Azure Data Factory, Prefect, or similar).

Qualified applicants must be authorized to work in the United States on a full-time basis. Lilly will not provide support for or sponsor work authorization or visas for this role, including but not limited to F-1 CPT, F-1 OPT, F-1 STEM OPT, J-1, H-1B, TN, O-1, E-3, H-1B1, or L-1.

Additional Preferences: Working knowledge of 21 CFR Part 11, ALCOA+, and GxP data integrity principles, or clear demonstrated ability to apply similar audit/compliance frameworks. Experience integrating data from LIMS, ELN, SDMS, or CDS systems (Benchling, LabVantage, OpenLABS, or equivalent).

Familiarity with pharmaceutical CMC data types: analytical results, batch records, stability studies, specifications. Experience with data mesh architecture or data product ownership models.

Knowledge of MLOps practices and preparing data for AI/ML model training in regulated environments. Exposure to regulatory submission data formats (eCTD, CTD, CDISC SEND/SDTM).

Experience with CI/CD pipelines (GitHub Actions, Azure DevOps) applied to data engineering workloads.