Neural Concept applies deep learning to engineering simulation, training geometric models that predict physical performance orders of magnitude faster than traditional solvers.
Founded in 2018 as an EPFL spin-off, it works with automotive and aerospace engineering teams.
