Marco Ciccone is a Postdoctoral Fellow at the Vector Institute, working with Prof. Colin Raffel. He obtained a Ph.D. from Polytechnic University of Milan and he interned at NVIDIA and NNAISENSE. His current research focuses on building modular intelligent systems that can efficiently learn to solve new problems by reusing, composing, and improving previously learned skills. This includes research on: Federated Learning, to enable collaboration and decentralization by sharing knowledge and resources while preserving privacy; Continual Learning, to enable continuous knowledge integration without forgetting; Modular Learning, for designing more reusable, maintainable, and efficient models. He served as Competition Track co-chair at NeurIPS in 2021, 2022 and 2023.