Data Engineer - Pharma R&D

At Roche you can show up as yourself, embraced for the unique qualities you bring. Our culture encourages personal expression, open dialogue, and genuine connections, where you are valued, accepted and respected for who you are, allowing you to thrive both personally and professionally. This is how we aim to prevent, stop and cure diseases and ensure everyone has access to healthcare today and for generations to come. Join Roche, where every voice matters.

Job Description:

The Data Engineer – Clinical Study Design sits at the intersection of data architecture, clinical science, and technology delivery, helping build the data foundation behind Study Designer, a digital product transforming how studies are designed. This role combines strong data engineering skills, an understanding of clinical study data and workflows, and technical curiosity to translate complex clinical data structures into reliable, scalable pipelines and models that power the platform's insights. The ideal candidate is curious, collaborative, AI-minded, and passionate about using data to enable smarter, faster, and more effective clinical study design.

Key Responsibilities

- Build ingestion pipelines for clinical trial protocols, ICF documents, SmPCs, CSRs, and published articles (PubMed, CTIS, ClinicalTrials.gov) - handling PDF parsing, text extraction, and structured data normalization

- Design and implement data models in Amazon Aurora (relational) and GraphDB (knowledge graph) to represent trial design entities: endpoints, eligibility criteria, study arms, interventions, therapeutic areas, and their relationships

- Develop embedding and vectorization pipelines to prepare extracted clinical text for RAG-based retrieval in LangGraph agentic workflows - chunking strategies, metadata enrichment, and vector store population

- Build and maintain ETL/ELT workflows that transform unstructured clinical content into queryable, linked data across both relational and graph stores

- Implement data quality validation specific to clinical data - protocol section classification accuracy, entity extraction completeness, cross-reference integrity (NCT IDs, EudraCT numbers, MeSH terms)

- Build data serving APIs (Python/FastAPI) that expose curated datasets to the Angular frontend and LangGraph agent layer

- Set up data lineage tracking and audit trails to support regulatory traceability of AI-generated trial design recommendations