In 2017 researchers and PhD student from the Physics Department and FBK took the deepPP initiative, focused on applications of Deep Neural Networks to to high energy physics and astrophysics. These powerful tools are used to faithfully reproduce the physics taking place at particle level, starting from the multitude of signals provided by gigantic detectors like ATLAS at the LHC or AMS on the International Space Station. Developing new tools and improving the interpretability of their predictions is currently the main objective of the deep initiative.