X-ray photoelectron spectroscopy (XPS) is a surface sensitive materials characterisation technique used to investigate material properties, including chemical composition, chemical depth distribution or core-shell nanoparticles. Extracting accurate quantitative information from XPS data requires typically trained experts to perform time consuming empirical peak fitting for each individual spectrum. In response to the growing demand for reliable, high- throughput quantitative interpretation, we present an automated analysis pipeline that combines the Simulation of Electron Spectra for Surface Analysis (SESSA) software with a transformer based neural network. In this study, SESSA was utilised to generate approximately 8.1 million spectra of 7,587 inorganic and organic bulk compounds. In combination with their corresponding chemical abundance labels, these data were utilised to train a supervised transformer to predict their chemical concentration. The present model has been demonstrated to correctly predict 90% of all compounds in the unseen test set, with a mean absolute error below 0.2 for most elements in the test data.