Andreas Paraskeva’s research revolves around the realm of parsimonious architectures, with a particular focus on the application of Automated Machine Learning (AutoML) methods for improving the efficiency of conversational agents and reducing their compute footprint. He is a member of the Automated Design of Algorithms (ADA) research group and is also involved in the LESSEN Project, a collaborative initiative that collaborates with a diverse team of researchers from various Dutch universities and industrial partners.