In recent decades, experimental and simulation-based high-throughput screening approaches have become increasingly important in the field of Materials Science. This trend is driven by the need to reduce the time and costs associated with conventional laboratory experimentation. In the domain of thin film technologies, particularly in the context of Physical Vapor Deposition (PVD), combinatorial sputter deposition has emerged as a robust experimental technique. Conversely, ab initio methods, such as Density Functional Theory (DFT), frequently encounter difficulties in reliably predicting critical properties, including mechanical behavior and the solubility of alloying elements. This challenge is especially relevant for the design of novel thin film materials, such as carbides, nitrides, and borides, where the extreme cooling rates inherent to PVD favor the formation of metastable and frequently unexplored solid solutions. Precise predictions in such non-equilibrium phase spaces would considerably benefit experimental optimization. However, a significant limitation of ab initio techniques is the substantial difference between the resulting phase diagrams and equilibrium ones typically obtained via CALPHAD. Over a decade ago, a model was developed that integrated combinatorial sputter deposition data with the insights derived from DFT and CALPHAD. This model took into account the effects of substrate temperature, target power, and residual film stress. According to literature reports, it achieved satisfactory predictive accuracy for metastable phase formation in both metallic (W1-xCux and V1-xCux [1]) and ceramic (Ti1-xAlxN and V1-xAlxN [2,3]) material systems. This study contributes to the existing methodology in two primary ways. Firstly, it eliminates the necessity for pre-existing CALPHAD databases. Secondly, it incorporates interfacial energy contributions arising from microstructural features such as preferred crystallographic orientation and grain morphology. The study's primary focus is on metastable ceramic thin films, and it commences with a comprehensive examination of the model's dependencies on experimental and DFT input data. This is followed by a reproduction and extension of the model for the metastable Ti1-xAlxN and V1-xAlxN systems, along with an outlook on ongoing efforts to apply the approach to well-studied carbide and boride systems. [1] Chang K., et al. Sci. Technol. Adv. Mater. 2016;17:210. [2] Liu S., et al. Acta Mater. 2019;165:615. [3] Liu S., et al. Acta Mater. 2020;196:313.