I develop statistical methods for evaluating, auditing, and monitoring AI systems in health care, including anytime-valid inference for deployed models, evaluation of large language models, and causal inference for hospital quality improvement.
I am a Staff Scientist and Data Scientist at UCSF working in Jean Feng’s lab and a member of the PROSPECT lab at Zuckerberg San Francisco General Hospital.
Previously, I was a postdoctoral researcher at the Regulation, Evaluation, and Governance Lab at Stanford Law School working on tax policy and algorithmic fairness.
I received my Ph.D. in Statistics from the University of Southern California’s Marshall School of Business department of Data Sciences and Operations where I was advised by Profs. Yingying Fan and Jinchi Lv.
Patrick Vossler
Selected Papers
- STAVAT: Stratified Anytime-Valid Atom Sampling for LLM Tier Placement STAI-X 2026 · Best Paper Award
- LLMs Judging LLMs: A Simplex Perspective AISTATS 2026
- Human-AI Co-design for Clinical Prediction Models npj Digital Medicine, 2026
- When the Domain Expert Has No Time and the LLM Developer Has No Clinical Expertise: Real-World Lessons from LLM Co-Design in a Safety-Net Hospital AAAI 2026 · Oral presentation
- Adaptive Auditing of AI Systems with Anytime-Valid Guarantees arXiv preprint, 2026
News
- Aug 2026STAVAT received the Best Paper Award at STAI-X 2026, the inaugural Conference on Statistics and Trustworthy AI, at Harvard.
- Jun 2026Human-AI co-design for clinical prediction models is out in npj Digital Medicine.
- May 2026LLMs Judging LLMs: A Simplex Perspective appeared at AISTATS 2026.
- Apr 2026Patient perspectives on AI decision support in a safety-net health system is published in JAMIA Open.
- Jan 2026Oral presentation at AAAI 2026 of our paper on LLM co-design with social workers in a safety-net hospital.