Associate Engineer, Technology

Abb Vie's mission is to discover and deliver innovative medicines and solutions that solve serious health issues today and address the medical challenges of tomorrow. Our focus spans immunology, oncology, neuroscience, and the Allergan Aesthetics portfolio.

We are seeking a technically versatile and self‑driven Associate Engineer, Technology I to join the Advanced Solutions Team within the DELOS (Data, Exploration & Linked Outcomes Solutions) Research Group, with a primary focus on leading the EXTRACT platform, an AI‑powered data engineering suite. This role combines data science, software engineering, and clinical informatics to automate the full lifecycle of data standardization (aligned to FDA/SDTM standards) and OMOP Common Data Model transformation.

The ideal candidate brings strong Python and SQL skills, hands‑on experience with large language models and RAG architectures, and a passion for building intelligent systems that replace manual data processes with automated, auditable AI pipelines executed directly against enterprise Impala infrastructure. The standardized data produced by EXTRACT directly supports the DELOS Patient Verse initiative, providing the clean, research‑ready data layer that downstream patient‑level analytics and insights depend on.

• Lead and execute agile sprints with stakeholders from all business domains, gathering requirements and delivering actionable data solutions.
• Harmonize and integrate patient‑level data (clinical trial, EHR/claims, etc.) across business lines.
• Partner closely with project owners to ensure data, tools, and AI solutions are scalable, fit‑for‑purpose, and impactful.
• Contribute to the organization’s long‑term data/AI/tool strategies by sharing hands‑on knowledge and workflow improvements.
• Drive engagement, adoption, and change management by actively collaborating with teams from early research.
• Design and build AI pipelines that ingest raw clinical trial data (Adverse Events, Lab Tests, Medical History, Procedures, Drug Names, Subject Exposure) and standardize it to FDA/SDTM regulatory formats using LLM‑powered term resolution, phonetic matching, and symbolic rule engines.
• Analyze and validate clinical trial datasets across large study libraries to evaluate and confirm their transformation into CDISC SDTM data standards.
• Architect and maintain a Retrieval‑Augmented Generation (RAG) system that indexes OHDSI clinical documentation (THEMIS, CDM field specifications, dbt‑synthea SQL patterns) into vector stores and injects relevant context into LLM inference for automated OMOP CDM field mapping.
• Build automated ETL SQL generation that reads source schemas via scan reports, produces Impala‑compatible INSERT/SELECT statements, and populates OMOP CDM tables replacing months of manual mapping with a single pipeline command.
• Develop vocabulary resolution systems that map clinical codes (e.g., LOINC, MedDRA, ICD‑10, RxNorm) to OMOP  using Athena vocabulary tables, embedding similarity, and LLM reasoning for ambiguous cases.
• Implement multi‑layer data quality validation (DQD constraint checks, Achilles descriptive analysis, Omop Checkout sanity checks) translated from R/JDBC to Impala SQL, producing automated HTML quality reports after every ETL run.
• Build neuro‑symbolic AI pipelines that route clinical term cleaning through four tiers — symbolic rules (YAML), phonetic algorithms, embedding similarity, and LLM inference — with full auditability and per‑tier traceability on every decision.
• Design and develop full‑stack applications (React frontend, Python‑based API backend, Impala database layer) including a mapping review dashboard with confidence scores, RAG citations, and approve/reject workflows for stakeholder sign‑off.
• Build and integrate interactive dashboards (Qlik Sense, Power BI) into applications to surface data‑quality, mapping, and stakeholder insights.
• Build semantic classification models that automatically determine the meaning, data type category, clinical domain, sensitivity level, and OMOP mapping target for every column in any source database.
• Develop LLM‑powered data profiling capabilities that analyze source…