Computational biologist · ML scientist

Arash Parsa

Clinical machine learning, bioinformatics, and the bench-to-computation bridge — based in the Bay Area.

I build predictive models and computational pipelines that turn complex biological and clinical data into decisions. My work sits where three worlds meet: applied machine learning, bioinformatics, and hands-on molecular biology. Right now that means clinical NLP on electronic health records, omics analysis, and cost-aware systems designed to actually run inside the constraints of a hospital.

I'm currently collaborating with Dr. Vivek Rudrapatna's clinical data science group at UCSF on a confidence-gated cascade for EHR phenotyping — a four-stage pipeline that keeps most inference on a cheap CPU-only path and escalates only the hard cases to a frontier model. Alongside it I work on clinical prediction, spatial transcriptomics, and biomedical ML with collaborators across UCSF and Stanford.

My background spans a BS in Bioengineering/Bioinformatics from UC Santa Cruz and an MS in Data Science from Georgia Tech, with prior work at Optelligence, Amgen, and ThermoFisher, and wet-lab roots in synthetic biology (iGEM).

Bay Area, California