We have successfully developed and trained a Machine-Learned Interatomic Potential (MLIP) for NbN, capable of accurately capturing the material’s structural and energetic behavior. This represents a significant milestone in our project. The next critical phase involves extending the application of this MLIP to systematically explore the effects of temperature, compositional variations (including different concentrations and configurations of Nb and N vacancies), and potentially different crystallographic phases of NbN. One key aspect that remains unexplored is the phonon dispersion of vacancy-containing NbN structures across a range of temperatures, which is essential for a comprehensive understanding of their dynamical stability. To achieve these goals and complete the theoretical foundation needed for publication, continued support and computational resources are necessary.