Jean-Benoit Delbrouck, PhD
Principal ML Scientist at HOPPR
Jean-Benoit Delbrouck is a medical AI researcher specializing in multimodal vision–language systems, medical foundation models, and clinically aligned evaluation, with publications as lead and senior author in top venues and journals including Nature Medicine, Nature Biomedical Engineering, ACL, EMNLP, NeurIPS, and ICLR, and over 2,000 citations. He is recognized for contributions to AI evaluation in radiology and medical imaging, including influential metrics and toolkits such as GREEN and RadGraph, as well as for the development of foundation vision–language models for medical imaging, including CheXagent.
Trained at Stanford University, he served as academic staff at the Stanford Center for Artificial Intelligence in Medicine and Imaging (AIMI), where he coordinated interdisciplinary research, led large-scale projects, mentored PhD, and contributed to widely adopted datasets, benchmarks, and evaluation frameworks for medical AI. He later joined Hugging Face as a Research Scientist, leading research on foundation models and evaluation, advancing open-source tooling, and building partnerships with medical and scientific communities to bridge academic research and industry-scale model development.
At HOPPR, Dr. Delbrouck currently serves as a Research Scientist, where his responsibilities include building high-quality training datasets, generating high-fidelity clinical labels from unstructured medical text, and leading the development of the vision–language model training framework for radiology and medical imaging applications.
