Published multiple papers on approximate inference, active learning, and their applications to sciences. Developed image superresolution and compression techniques at Magic Pony Technology. Worked on various ML projects, including recommender systems and fair machine learning.
Research Experience
Worked in various jobs in the London tech/startup sector after his PhD. Joined Magic Pony Technology, where he developed deep learning-based image superresolution and compression techniques. After Twitter's acquisition of Magic Pony, worked on a range of ML topics, including recommender systems and fair machine learning.
Education
PhD in Bayesian machine learning from the Engineering Department at the University of Cambridge, supervised by Carl Rasmussen, Máté Lengyel, and Zoubin Ghahramani. Research topics included approximate inference, active learning, and their applications to sciences.
Background
Associate Professor of Machine Learning at the University of Cambridge, interested in principled deep learning techniques such as optimization, generalization, representation, transfer, and meta-learning. Focuses more on understanding than developing new techniques.
Miscellany
Runs a personal blog called inFERENCe, sharing insights on machine learning, statistics, and current research topics.