The mission of the ELLIS unit Cambridge is to build on the excellent machine learning and AI infrastructure already available within the University of Cambridge and serve as a stepping stone towards creating a center of excellence. Many of the members of the ELLIS unit Cambridge are strong in Bayesian statistics and probabilistic machine learning. These types of methods are expected to play a key role in addressing some of the limitations of existing approaches: lack of robustness, data-efficiency, uncertainty awareness, flexible adaptation and understanding causality. Other unit members work on specific application areas: language modeling, healthcare, computer systems and molecular modeling. The significant strength in probabilistic and Bayesian machine learning makes the ELLIS unit Cambridge unique in Europe and a valuable asset for the ELLIS society. Furthermore, the joint methodology used by the members of the unit, together with a diverse portfolio of expertise across different application areas, creates an excellent foundation for broad impact.