Xuguang Ai

my_profile.jpeg

Regenstrief Institute

Indianapolis, IN 46202

I am a data scientist at Regenstrief Institute in Indianapolis, IN. Previously, I was a data scientist & software engineer in the Department of Biomedical Informatics & Data Science at Yale University, where I had the great privilege of working with Qingyu Chen and Hua Xu. I hold a Master’s degree in Physics and a Master’s degree in Data Science from the University of Kentucky, where I was fortunate to be advised by Ramakanth Kavuluru during my second Master’s.

My research focuses on advancing AI in healthcare, particularly in:

  • End-to-end information extraction
  • Ophthalmology-specific (multimodal) LLMs
  • RAG and agent systems in healthcare
  • AI development leveraging real-world data (RWD) and real-world evidence (RWE), such as the OMOP Common Data Model

news

Dec 26, 2025 Our paper Benchmarking LLMs for Ophthalmology (BELO): An Expert-Curated Dataset and Evaluation Framework for Knowledge and Reasoning accepted at Ophthalmology Science is now online!
Oct 23, 2025 Excited to share that our paper AgentMD: Empowering Language Agents for Risk Prediction with Large-Scale Clinical Tool Learning has been published in Nature Communications!
Jul 31, 2025 Excited to share that our paper Ophthalmological Question Answering and Reasoning Using OpenAI o1 vs Other Large Language Models has been published in JAMA Ophthalmology!

selected publications

  1. Nature Comms
    Agentmd: Empowering language agents for risk prediction with large-scale clinical tool learning
    Qiao Jin, Zhizheng Wang, Yifan Yang, and 8 more authors
    Nature Communications, 2025
  2. JAMA Ophthalmology
    Ophthalmological question answering and reasoning using OpenAI o1 vs other large language models
    Sahana Srinivasan, Xuguang Ai, Minjie Zou, and 8 more authors
    JAMA ophthalmology, 2025
  3. Nature Comms
    Benchmarking large language models for biomedical natural language processing applications and recommendations
    Qingyu Chen, Yan Hu, Xueqing Peng, and 8 more authors
    Nature communications, 2025