Thin film photovoltaic (PV) devices are multi-component and multi-layer complex structures composed of an elevated number of micro- and nano-layers of different materials. As such, their performance is controlled by a large number of complexly intertwined variables associated to the characteristics of each layer and interface. Nevertheless, the main techniques for monitoring such complex systems remain to be limited to compositional and/or J-V analysis. The limitations of these techniques are one of the main bottlenecks for high-throughput development of novel materials and devices. Spectroscopic characterization techniques, such as Raman and photoluminescence (PL), are becoming widespread for advanced material characterization due to their easy application and highly informative output about different materials properties in a fast and non-destructive manner. Furthermore, the application of Artificial Intelligence for the analysis of spectroscopic data represents a powerful step forward that allows dealing with the high-dimensionality data generated from the combinatorial analysis of highly complex systems like thin film PV devices. However, in order to implement this type of advanced analyses effectively, appropriate methodologies need to be developed. In this work, we present a methodology based on the combination of spectroscopic techniques and Machine Learning (ML) for the fine monitoring of the quality of complex absorber layers as well as for the accurate prediction of optoelectronic parameters of thin film solar cells.