His current research focuses on the interplay of four principles of statistics and data sciences. They include heterogeneity of data, interpretability of models, stability, and scalability of optimization and sampling algorithms.