In the Materials Informatics Lab, or the MI-Lab, our research revolves around the fusion of data science (AI or machine learning) with material property computations, in order to guide experiments and accelerate the design and discovery of advanced materials. We primarily utilize first-principles electronic structure methods (DFT), molecular dynamics simulations (MD), and data-driven techniques, including deep learning, multi-fidelity learning and reinforcement learning. We focus on the synthesis of shape-memory alloys, discovery of “green” polymers, and automation of material production laboratories.