I will discuss two projects I recently led involving supermassive black holes (SMBHs) with relatively light masses (1e4-1e6 Msun). Transient events, like tidal disruption events (TDEs), occurring on these objects allow us to study accretion processes on human-approachable time scales, and the relation between these SMBHs and their host galaxies could potentially inform us about the formation of primordial black holes and their growth mechanisms. I will present the properties of a sample of X-ray-selected active galactic nuclei (AGN) in local dwarf galaxies, compiled by exploiting recently released eROSITA data. I will demonstrate how this sample can teach us about the high-redshift counterparts of the SMBHs powering this local AGN. I will also show the result of a data mining of the Chandra catalog of X-ray sources, which led to the identification of an overlooked TDE. This search was performed both via a manual procedure and a novel machine-learning algorithm, and I will illustrate the advantages and potential of this novel approach.
Not-so-supermassive black holes: where and why to find them!