Gas chromatography – mass spectrometry (GC-MS) is an important tool in contemporary untargeted chemical analysis, where the batch analysis of sample series and subsequent generation of peak tables are still commonly subject to software-uncertainty leading to issues in reproducibility and hypothesis testing. Using tensor-based modelling in combination with other machine learning tools, we were able to provide a completely automated method for turning GC-MS data into a peak-table that is absent of user-interactions, avoiding user induced differences in the peak tables. The developed tools are integrated into the software package called PARADISe. The results of using the fully automated version of PARADISe are illustrated using experimental GC-MS data. The presented approach still has room for improvement, especially when the data collinearity is broken, such as in the case of peak saturation. The proposed automated approach provides marked improvements over current analysis, including but not limited to the analysis time and reproducibility.