cv

Software Engineer

Qualcomm Machine Learning Group

Worked across Qualcomm's inference engine stack.

• Op Kernels in OpenCL/Hexagon
• Model Efficiency techniques like quantization
• Caching framework to optimize model load times

MS

University of Wisconsin-Madison

Machine Learning

BS in Electrical and Computer Engineering

University of Wisconsin-Madison

Computer Architecture

Dean's List (6 out of 8 semesters)

College of Engineering

Biophysics-based protein language models for protein engineering

Nature Methods

Protein Language Models trained on Biophysical Simulations.

Exploring zero-shot structure-based protein fitness prediction

GEM Workshop, ICLR '25

Evaluating Protein fitness predictors using structure data. Disected performance on AlphaFold predicted vs experimental structures, disordered vs ordered regions.