The article presents a description of the developed software platform that provides automation of data storage server state monitoring using a hybrid algorithm. The developed algorithm combines an ontological approach and machine learning elements. A comparative assessment of various approaches was carried out: cluster analysis, autoencoders, LSTM, fuzzy ontologies and hybrid models - when analyzing methods for searching for anomalies in time series. The hybrid evaluation model showed the best results for all criteria. It was chosen to develop a prototype of the software platform. Experiments were conducted on hardware storage system data. The software platform proved its effectiveness in solving the problem.