Education
All three degrees at the University of Washington — moving from undergraduate research straight into a PhD focused on trustworthy machine learning.
Awards & Honors
Awards and fellowships spanning research, leadership, and national data science competition wins.
Publications
Peer-reviewed work at the intersection of large language models, privacy, and health informatics.
Professional Experience
Industry and research roles across Microsoft, Zillow, the FDA, and the University of Washington.
- Joining Zillow's applied science organization to work on a new privacy-focused project.
- Working on self-improving agentic systems and their evaluation.
- Worked on synthetic data generation using differential privacy and large language models under the supervision of Dr. Habernal.
- Worked on agentic AI systems to enhance real estate technology and customer experience.
NIH AIM-AHEAD Research Fellowship on Acute Myeloid Leukemia subtyping
- Improved scGPT's cell subtype classification on genomics and bulk seqRNA data for AML, optimizing for fairness and accuracy across demographic groups.
LLM-assisted Tuberculosis Treatment Adherence Tool
- Led an interdisciplinary team training an LLM for high performance in multilingual settings while preserving the privacy of patient dialogues.
Caring For Caregivers Online
- Led an interdisciplinary team training an LLM chatbot to deliver parts of Problem-Solving Therapy to assist family caregivers.
- Evaluated a medical synthetic image generator by designing a downstream segmentation and detection-based evaluation pipeline in PyTorch.
- Created a FHIR-specific migration system package managing store updates.
- Refactored a predictive language model into a separate containerized service.
- Won a NASA-sponsored competition aimed at improving flight departure time prediction with tree-based ML models.
- Designed the winning model with feature engineering and supervised learning, improving accuracy over the baseline by 50%.
- Adapted the model to the federated learning paradigm via the PyTorch Flower framework.
Leadership
- Communicated with graduate CSS students to understand their questions and encapsulate the overall concerns for the CSS program.
- Presented before CSS faculty to represent current students' academic needs and contribute to the regular CSS program meetings.