Abstract: A system can identify a group of computer services that correspond to respective computer hardware. The system can extract features that respectively identify characteristics of the computer services, a feature of the features comprising ratio information representative of a ratio of a first price of a computer service of the group of computer services to a second price of a corresponding computer hardware of the respective computer hardware that corresponds to the computer service. The system can convert the features into a numerical representation. The system can cluster the computer services into multiple clusters based on the numerical representation of the features. The system can identify, from a cluster of the multiple clusters and based on a machine learning model, an anomalous price for at least one computer service of the group of computer services that belongs to the cluster. The system can store an indication of the anomalous price.