In the modern world, information systems it’s important tool in corporate structures and business processes. ERP systems are widely used for automating business processes and improving communication between employees in companies. However, like any other system, ERP systems can vary in quality, which in turn affects the efficiency and feasibility of business processes. This article describes a development an information system that uses a neural network to analyze the quality of ERP systems. The input parameters of the neural network are a subset of 12 quality parameters defined in ISO 25010. The developed neural network is based on a multilayer perceptron architecture. The newly created reduction methodology using the mean of L1 and L2 regularization allowed reducing the connections in the neural network and increasing the model accuracy. As a result, the neural network has a high accuracy, sufficient for practical use. A web application has been developed for the neural network, which is a graphical user interface that allows interacting with the neural network and evaluating the quality of ERP systems. Thus, the created information system is capable of significantly simplifying and speeding up the process of analyzing the quality of ERP systems.