Dani Kiyasseh
Building large-scale vision models to understand surgery from video, so clinicians get better insights.
Research
Publications in ICML, NeurIPS, Nature Biomedical Engineering, and Nature Communications on surgical AI and clinical machine learning.
→Experience
From Oxford PhD to Caltech postdoc to founder — across academia, industry, and startups including Mayo Clinic, Flatiron Health, and Vicarious Surgical.
→Writing & Blog
Long-form posts on surgical AI, clinical deep learning, and the ideas behind the research — written for scientists and builders alike.
→Decoding surgeon activity from video.
A vision transformer that tracks surgical activity from operating room videos — featured on the front cover of Nature Biomedical Engineering. Developed during my postdoctoral fellowship at Caltech.
Read the paper →
From Oxford thesis to founder.
I have been operating at the intersection of AI and healthcare for the past 8 years. My recent focus has been on building large-scale vision-based foundation models to better understand surgery — a mission we started at Halsted AI, a Techstars-backed company I founded in 2025.
I completed my postdoctoral fellowship at Caltech, developing AI systems to track surgeon activity from videos. Before that, I received my PhD from the University of Oxford, where I developed deep learning models for limited labeled clinical data. I graduated from Johns Hopkins with a degree in biomedical engineering.
I was named to the Forbes 30 Under 30 and MIT Innovators Under 35 in the MENA region, and selected as a Rising Star in Engineering in Health by Columbia, Cornell, and Johns Hopkins. I have conducted AI research at Ford Motor Company, the Mayo Clinic, Merck, Flatiron Health, and Vicarious Surgical.
Selected work.
Across research, recognition, and building.
Let's connect.
Book a 1-hour consulting session on end-to-end AI development — from idea to production. Or reach out about surgical AI, research collaborations, and building in health.