In this work, we describe a project, jointly started by DEIS University of Bologna and Dianoema S.p.A., in order to build a system which is able to monitor nosocomial infections. To this purpose, the system computes various statistics that are based on the count of patient infections over a period of time. The precise count of patient infections needs a precise definition of bacterial strains that is found by applying clustering to data on past infections. Moreover, the system is able to identify critical situations for a single patient (e.g., unexpected antibiotic resistance of a bacterium) or for hospital units (e.g., contagion events) and alarm the microbiologist.