The core problem in single-cell mass spectrometry-based proteomics is low signal-to-noise ratio in the MS spectra, which makes it difficult to discriminate signal from background noise, requiring much more performant data analysis. For this, typical analysis pipelines rely on peptide libraries, which contain only peptides already known to occur in a given cell type, limiting the possible set of identifiable proteins. For prokaryotes, this strategy is less effective: the proteomes of these organisms are not known in detail, and are typically more dynamic when faced with external triggers. In my PhD project, I will develop alternative data analysis approaches that enable single-cell mass spectrometry-based proteomics for prokaryotic cells without peptide libraries. These approaches include peptide behaviour predictors (e.g. retention time, ion mobility separation) and peptide identification algorithms that are optimised for instruments and data acquisition methods applied in single-cell proteomics workflows.