Levi Kaster Bio
I am a 2nd-year PhD student in Biomedical Informatics and Data Science at Washington University School of Medicine in St. Louis, within the Institute for Informatics, Data Science, and Biostatistics. My research focuses on leveraging large language models and rule-based systems for clinical entity extraction and improving patient care. I am interested in the safe and effective deployment of AI in healthcare.
Current projects include predictive modeling for Neurofibromatosis Type 1 (NF1), developing LLM prompting methodologies for clinical information extraction (e.g., retrieval-augmented generation and divide-and-conquer approaches), and advancing the transfer of interoperable NLP technologies across academic medical centers.
I graduated with a B.S. in Data Science from the WashU McKelvey School of Engineering in December 2023, and joined my PhD program the next year. I am co-advised by my amazing advisors Drs. Philip R. O. Payne and Aditi Gupta.
Selected News
2025-09-12 — Our paper comparing rule and large-language model based entity extraction methods for NF1 was published in JAMIA!
2025-05-31 — Our team won a $5,000 prize for our project at the CAIDF Hackathon. We developed multi-institutional 30-day fall readmission risk models that incorporated LLM-extracted phenotypes from clinical notes, and displayed the risk with an interactive dashboard that incorporated patient level LLM-generated summaries, social determinants of health, and the top predictive features.
2024-11-12 — Gave a talk at the 2024 Child Neurology Society Meeting on utilizing LLMs to extract functional biomarkers from multi-institutional clinical notes.