The goal of the project is to provide computational resources to the FunMat-II Competence Center, the FurCoat consortium, and our activities within SeRC. The scientific goal is to establish a high-accuracy computational platform based on ab initio molecular dynamics simulations combined with active machine learning of interatomic force fields. This platform will be used to investigate the high-temperature thermodynamic and mechanical properties of materials and to generate data for mesoscopic models. The application directions are defined in close collaboration with our industrial partners, including Sandvik Coromant, Seco Tools, and Westinghouse Electric AB. Our aim is to build the unique Hard Coating Alloy Database (HADB), supporting multiscale investigations of materials for hard-coating and nuclear applications. We primarily focus on refractory B1, B3, and B4 nitrides, carbides, and borides, together with bcc high-entropy metallic alloys, for applications in the hard-coating industry and nuclear reactors. The project has three major objectives: (i) to investigate the microscopic mechanical properties of refractory nitrides and nitride multilayers using machine-learned interatomic potentials; (ii) to generate thermodynamic data for Nb-, Ta-, Cr-, and Si-containing alloys; and (iii) to utilize neural-network-based approaches to learn, explore, and predict the corrosion properties of alloys within HADB. In parallel, we use the full-potential local-orbital (FPLO) software to analyze the topological features of the electronic structure of metastable and other exotic material phases. Through the FunMat-II Competence Center, we are in close collaboration with Seco Tools and Sandvik Coromant. We also collaborate with world-leading experimentalists in the field of high-pressure physics. Our research is supported by the Swedish strategic FunMat-II and FurCoat initiatives, as well as the Interdisciplinary Laboratory for Advanced Functional Materials (AFM) at Linköping University. The ab initio calculations will be performed using VASP and Quantum ESPRESSO (QE), while classical molecular dynamics simulations will be carried out using LAMMPS.