The current approach to designing electric machines is carried out using a 'cascade' algorithm, wherein sequential calculations are performed according to formulas derived from electromagnetic theory. However, this approach is tailored to each specific type of electric machine, making modification of projects difficult and not always feasible. One promising solution to address this issue is to leverage HPC (High-Performance Computing) calculations in conjunction with machine learning capabilities. This chapter examines established methodologies for designing and optimizing electric machines, ultimately advocating for the integration of machine learning and modern optimization algorithms.