Before coming to Sandtable, Bernard held a two-year post-doc position at Queen Mary University, studying the formation of dyads in online social networks and running labs for the Social Networks and Coordination & Social Dynamics courses. Prior to this, he completed a PhD at Brunel University in complex networks (thesis title: “The Structure and Dynamics of Complex Systems”).
Bernard joined Sandtable in 2010, bringing his complex networks theory expertise into agent-based simulations. His role has evolved over time, and as Senior Data Scientist he now works on designing and implementing agent-based models (C++, Python, Cython), conducting EDA, and working with our big data clusters. He has worked on most of the projects that have passed through Sandtable. He is currently building a simulation of the UK grocery market, a project that is challenging both because of the volume and variety of the data it uses, the compute power required to run the model, and the need to provide compelling visualisation of the model outputs. Agent-based modelling is a good fit for Bernard’s holistic approach to model building. Bernard is particularly proud of his work on the smoking simulation he built for Public Health England, as it has the potential to save and improve lives. He particularly enjoys mastering new technologies, and has recently become Sandtable’s go-to man for questions about performant Python (Cython).