Pranav Rajpurkar

Pranav Rajpurkar is a computer scientist and Associate Professor of Biomedical Informatics at Harvard Medical School. He co-founded a2z Radiology AI. He received his BS, MS, and PhD degrees in computer science from Stanford University.

He leads the Rajpurkar Lab, which develops and evaluates AI systems for medical imaging, multimodal clinical data, clinical reasoning, and clinician–AI collaboration. His research asks how these systems can move beyond benchmark performance to become reliable and useful in healthcare.

Pranav Rajpurkar is a computer scientist and Associate Professor of Biomedical Informatics at Harvard Medical School. He co-founded a2z Radiology AI. He received his BS, MS, and PhD degrees in computer science from Stanford University.

At Harvard, the Rajpurkar Lab develops models, datasets, evaluation methods, and simulation environments for medical AI. Current work spans multimodal interpretation of medical images and text, clinical reasoning agents, clinician–AI collaboration, and systems that connect perception to action.

Rajpurkar has authored more than 200 research publications, which have received 58,000+ citations according to Google Scholar. His work appears in Nature, Nature Medicine, Cell, The Lancet, the New England Journal of Medicine, and other medical and computing venues. He also co-authored The Robot Doctor Will See You Now with Eric Topol for The New York Times.

He co-founded a2z Radiology AI. Its first product, a2z-Unified-Triage, is FDA-cleared to prioritize adult abdomen and pelvis CT studies with suspected findings across seven conditions.

Pranav Rajpurkar is a computer scientist and Associate Professor of Biomedical Informatics at Harvard Medical School. He co-founded a2z Radiology AI. He received his BS, MS, and PhD degrees in computer science from Stanford University.

The Rajpurkar Lab studies how AI systems can interpret medical data, reason through clinical tasks, and be evaluated under conditions that better reflect practice. The group builds multimodal models, datasets, benchmarks, clinical simulation environments, and methods for clinician–AI collaboration across imaging, text, video, and physiological signals.

Recent themes include generalist medical imaging models, dynamic evaluation of clinical agents, grounded report generation and assessment, coordinated AI systems, and early work on procedural and embodied medical AI. The goal is to measure what these systems can and cannot do, then develop methods that make them more reliable and useful in healthcare.

Rajpurkar has authored more than 200 research publications, which have received 58,000+ citations according to Google Scholar. His work appears in Nature, Nature Medicine, Cell, The Lancet, the New England Journal of Medicine, and other medical and computing venues. He also co-authored The Robot Doctor Will See You Now with Eric Topol for The New York Times.

Rajpurkar co-founded a2z Radiology AI. The company’s FDA-cleared a2z-Unified-Triage analyzes adult abdomen and pelvis CT studies and flags suspected cases across seven conditions for worklist prioritization.

Recognition includes MIT Technology Review’s Innovators Under 35, Forbes 30 Under 30 in Science, Nature Medicine’s early-career researchers to watch, and a 2026 ACL Test-of-Time Paper Award for the 2016 SQuAD paper. Rajpurkar is an instructor for the three-course AI for Medicine series, whose courses have reached 90,000+ learners on Coursera.