Using electronic health record (EHR) data to answer clinical research questions is challenging. It requires understanding the research question, finding relevant information across different sources, and combining that information to produce meaningful results. We developed an agentic large language model (LLM) pipeline that decomposes investigator-authored requests into source-specific tasks, generates and executes SQL on EHR, integrates structured and unstructured EHR evidence, and applies study-specific clinical note adjudication.
Agentic Pipeline for Large-Scale Clinical Data Queries in Enterprise Electronic Health Records
calendar_today
November 6, 2026
person
Pawar, K., Tavabi, N., Kalepalli, A., Kalepalli, N., Blackmon, M., Williamson, A., Kiapour, A.
domain
medrxiv