Data Architect, Data Foundry

Overview

Lilly Small Molecule Discovery is purpose-built to create molecules that make life better for people. Data Foundry is a multidisciplinary team within Discovery Technology and Platforms (DTP) that enables AI-native drug discovery through four integrated pillars: Architecture4Insight (data infrastructure and scientific software), Methods4Insight (analytical and computational methods), Automation & Scale4Insight (lab automation and agentic workflows), and Preparedness4Insight (data governance and readiness).

Responsibilities

• Design and implement data models, schemas, and ontologies for chemical, biological, and automation-generated data that serve discovery workflows across the portfolio.

• Define and maintain controlled vocabularies, metadata standards, and FAIR-compliant data frameworks in partnership with Preparedness4Insight.

• Implement semantic data standards (RDF, OWL, SPARQL) and ontology engineering practices to create interoperable, machine-readable scientific data.

• Design and implement data lakehouse architecture using modern platforms (Databricks, Snowflake, or equivalent), including data storage patterns, partitioning strategies, and query optimization.

• Build and optimize ETL/ELT pipelines using Spark, dbt, or similar tools to transform raw scientific data into analytical and ML-ready formats.

• Implement real-time and streaming data integration (Kafka, Kinesis, event-driven patterns) connecting LIMS, instruments, and lab automation systems to the data infrastructure.

• Design and implement knowledge graphs (Neo4j, Amazon Neptune, TigerGraph) that capture molecular, target, pathway, and experimental relationships across the discovery landscape.

• Architect specialized data solutions: array databases (TileDB) for genomics/imaging, document stores (MongoDB) for experimental records, and vector databases for embedding-based retrieval supporting ML and RAG workflows.

• Build query and traversal patterns that enable scientists and AI agents to ask relational questions across the entire data landscape.

• Partner with scientific software engineers to ensure data architectures are implementable, performant, and well-documented.

• Collaborate with Methods4Insight to design data structures that support analytical model training, deployment, and evaluation.

• Work with Tech@Lilly to define scaling strategies, ensure enterprise compliance, and transition data architectures to production-grade management.

• Contribute to build-versus-buy-versus-adopt decisions by evaluating commercial and open-source data platforms against Data Foundry requirements.

Basic Requirements

• B.S. or M.S. in Computer Science, Data Science, Bioinformatics, Computational Biology, Information Science, or related STEM field; Ph.D. valued for ontology and knowledge graph roles.

• SQL skills and experience in multiple database paradigms (relational, graph, document, columnar, key-value).

Preferred Qualifications

• Expertise in at least one of: data modeling/ontologies, data platform engineering (Databricks, Snowflake, Spark), or graph/specialized databases (Neo4j, Neptune, MongoDB).

• Familiarity with cloud platforms (AWS, Azure, or GCP) and modern data integration patterns.

• Understanding of scientific data types and experimental workflows in life sciences or pharma (chemical, biological, HTE data).

• Strong communication skills with ability to translate data architecture concepts for both technical and scientific audiences.

• Pharmaceutical or biotech research industry experience, particularly in discovery data management or research informatics.

• Experience with semantic web technologies: RDF, OWL, SPARQL, Protégé, or equivalent ontology engineering tools.

• Hands-on experience with graph databases (Neo4j, Neptune, TigerGraph) and knowledge graph design patterns for scientific data.

• Data lakehouse architecture experience: Databricks (Delta Lake, Unity Catalog), Snowflake, or equivalent; ETL/ELT with Spark, dbt.

• Experience with streaming/real-time data platforms (Kafka, Kinesis, Flink) and event-driven architectures.

• Familiarity with LIMS, ELN systems (e.g., Benchling), and laboratory instrument data integration.

• Experience with vector databases (Pinecone, Weaviate, pgvector) and embedding-based retrieval for ML/RAG applications.

• Array database experience (TileDB, Zarr) for genomics, imaging, or high-dimensional scientific data.

• Experience with bioinformatics data formats (FASTA, BAM/CRAM, VCF) and biological sequence databases; familiarity with NGS data pipelines and proteomics data management.

• FAIR data principles implementation experience and Data Readiness Level frameworks.

• Scientific data standards and controlled vocabularies in chemistry (InChI, SMILES) or biology (Gene Ontology, UniProt, pathway databases such as Reactome or KEGG).