Sheroze is a graduate student in Computational Science at the Oden Institute at the University of Texas at Austin. He is currently researching large-scale Bayesian inverse problems and deep learning models applied to physical systems.

His thesis is on scalable algorithms for data-driven statistical inference of parameters in nonlinear dynamical systems. His most recent work combines dimensionality reduction, physics-informed deep learning models, and Bayesian inference to not only efficiently infer high-dimensional parameters given observational data but also quantify their uncertainty: a critical aspect of robust inference algorithms.