My interests are in Gaussian process and other nonparametric models, in variational inference for graphical models, and in large scale problems. I did some work in semi-supervised learning as well. Recently, I got interested in models for neuronal spike data, in gene regulatory network identification (systems biology), in natural image modelling, and in other applications of sparse Bayesian models. Very recently, I began working on undersampled MRI reconstruction and design of MRI sampling trajectories using Bayesian methods. I am also the chief (and only) developer of the LHOTSE toolbox for adaptive statistical models, which you may find useful in your next high-performance scientific C++ application.