The aim of the BITMAP project "aB-IniTio calculations and MAchine learning for suPerconducting collective phenomena in novel materials" is to propose a workflow based on the combination of realistic Density Functional Theory (DFT) calculations with the Renormalization Group (RG) approach to superconducting Fermi surface instabilities. The latter is based on the pioneering work of Kohn-Luttinger where one can integrate out the high energy degrees of freedom perturbatively, and obtain effective attractive BCS interactions in non-s-wave channels. Once the superconducting pairing is known, as encoded in the superconducting gap function, a machine learning-based diagnostic procedure of the topological properties will be performed, upon the creation of specific ad-hoc convolutional neural networks. The project will allow the experience researcher to merge his present skills in the computational modeling of complex materials with modern concepts of machine learning, a sector that nowadays is expanding fast enough to easily foresee its applications in everyday life.