Application of machine learning for automated assessment of business process management in IT enterprises

Abstract

In the modern conditions of the digital economy, IT enterprises are faced with rapid changes that require rapid response and innovative approaches to business process management. The growth of data volumes and competition emphasize the need for flexibility, accuracy and efficiency of solutions. Optimization of business processes through automation and intelligent technologies, in particular machine learning, is becoming a key factor in the sustainable development of IT companies in conditions of uncertainty, contributing to increased competitiveness. The goal is to analyze and evaluate the possibilities of using machine learning for automated assessment of the quality of business processes in IT enterprises. The tasks include the study of ML methods in business analytics, the integration of cognitive models, the development of a structural forecasting model and a matrix of algorithm applications.

The article uses the analytical generalization method to synthesize data on machine learning in IT process management, which allowed us to identify key patterns. The comping method was used to structure information into a matrix of algorithm application, ensuring a logical organization of results. System search was used to collect relevant sources and empirical data, which contributed to a comprehensive analysis and substantiation of conclusions regarding business process automation. A matrix of ML algorithm application for forecasting, resource planning, and quality control in IT enterprises was developed. A cognitively oriented M-network model was proposed, which increases the accuracy of process assessment. The dependence of efficiency on data quality and the level of expertise was revealed.

Further research is possible in the direction of developing explainable artificial intelligence technologies to increase the transparency of machine learning solutions in IT process management. The implementation of edge learning for real-time data processing on IoT devices is promising, which will accelerate decision-making. Integration of AutoML tools can expand the availability of technologies for small IT companies. It is also advisable to improve cognitive models, in particular M-networks, taking into account the ethical and legal aspects of data processing. Research into the adaptability of models to non-standard scenarios and improving the quality of input data will contribute to the effectiveness of automated assessment of business processes in the digital environment.