I am a mathematician interested in designing more efficient and capable learning algorithms. I am building Untoken, an AI research company currently in stealth.

Previously, I was an Eric and Wendy Schmidt Fellow at the Broad Institute of MIT
and Harvard, where I developed data infrastructure, models, and evaluation
frameworks motivated by problems in single-cell biology.

I received my PhD from the Institute of Science and Technology Austria (ISTA),
where I worked on computational topology and geometry with Herbert
Edelsbrunner.

NEWS:
• May 2026: New preprint out!! We introduce a hierarchy-aware loss function that consistently improves image classification across architectures and datasets. 
• May 2026: scDataset was accepted at ICML, looking forward to presenting it in Seoul in July!
• Jan 2026: Our paper introducing hierarchical cross-entropy for cell-type annotation is now published in Nature Computational Science!