The COVID-19 pandemic caught the world unprepared. Vaccines and monoclonal antibodies (mAbs), developed in record time, mitigated the health and economic damages, however our reaction has always been one step behind the virus evolution, and emerging variants repeatedly escaped our interventions. The omicron variant escaped humoral immunity generated by most vaccines and mAbs by mutating immunodominant epitopes. The extremely potent mAb developed by our laboratory also lost potency against omicron. Here we propose to develop vaccines and monoclonals neutralizing existing and future variants of SARS-CoV-2 and other coronaviruses, by targeting immunologically subdominant regions which are less susceptible to antigenic variation. We will isolate mAbs from individuals who had infection and multiple vaccinations, whose repertoire is enriched in B cells encoding broadly neutralizing antibodies, to build a map of the broadly shared epitopes. Structural prediction and clustering of the immune repertoire through deep neural networks will be used to improve the breadth of coverage of the mAbs. The Monte Carlo-based sequence design of Rosetta and free energy perturbation calculations will be used to in-silico “design protein-binding proteins” and identify newly designed immunogens which can be loaded on nanoparticles and used as vaccines. This approach will provide broadly protective mAbs and vaccines proactively designed to neutralize all variants of SARS-CoV-2 and new coronaviruses that are very likely to jump from animals to humans in the future. If successful, the approach to map subdominant epitopes and use of genomic and structural information to design mAbs and vaccines targeting subdominant, broadly conserved epitopes, will pave the way to approach other pathogens with high antigenic variability such as influenza and HIV viruses, Plasmodium spp. and antibiotic resistant bacteria. This will strongly increase European competitiveness in fighting infections.