Location: New York, NY
Cap-Exempt H-1B Position — No lottery required
We are seeking an outstanding postdoctoral researcher to join a collaborative initiative at the intersection of artificial intelligence and clinical medicine. Based in the Departments of Pathology and Medical Physics, this position offers a rare opportunity to develop AI algorithms — including cutting-edge large language models (LLMs) and deep learning architectures — that will be directly integrated into clinical workflows at one of the world's premier cancer centers. This is not a purely theoretical research position. The tools you build will be deployed to support real clinical decision-making, with measurable impact on diagnostic accuracy, workflow efficiency, and ultimately patient outcomes. You will work at every stage of the pipeline — from algorithm design and model development to clinical validation and implementation — ensuring your research translates from bench to bedside. This position has a strong publication trajectory: completing the project is expected to result in multiple high-impact publications in both technical AI and clinical journals, giving you visibility across both communities. **Responsibilities** You will design, develop, and validate novel AI methods — spanning large language models, computer vision, and multimodal learning — tailored to solve real-world challenges in oncology and pathology. Working closely with clinicians, you will identify workflow bottlenecks, build intelligent tools to address them, and rigorously evaluate their clinical utility. Your work will directly inform how physicians interpret data, make treatment decisions, and manage patient care. **Qualifications** You hold a PhD in biomedical engineering, computer science, physics, mathematics, or a related quantitative discipline. You bring strong programming skills in Python and hands-on experience with modern deep learning frameworks. Familiarity with machine learning, computer vision, natural language processing, or medical image analysis is highly preferred. Above a