Binary neutron star (BNS) mergers are thought to be the dominant, or one of the dominant, production sites of rapid-neutron capture (r-process) elements in the Universe. Approximately half of all elements heavier than iron are synthesised via this process. Despite the r-process being extremely important for heavy element production, there is much we still do not understand about this production mechanism in BNS mergers. To date, we have observed one confirmed optical counterpart to a BNS merger - the kilonova (KN), AT2017gfo. Since its discovery in 2017, there has been several attempts to directly identify different r-process species in the observed spectra (strontium, caesium and tellurium). This proves to be difficult, due to the lack of complete atomic data for the heavy elements that would be of interest in this type of study. Additionally, the works to date have not considered realistic compositions from theoretical nucleosynthesis calculations. In this talk, I will present the results of our modelling of AT2017gfo, with compositions obtained from such a calculation. From this, we can make more robust predictions about the composition of the ejecta. Additionally, I will present the results of an analysis we performed with a new atomic dataset we generated for neutral, singly- and doubly-ionised platinum and gold (both 3rd r-peak elements). From these types of study, we can attempt to directly identify features corresponding to individual species. From this, we can infer properties about the KN ejecta, such as electron fraction, which helps to inform us about which r-peak is favoured in the elemental synthesis in KN ejecta.