ORGANIZATION OF INFORMATION FLOWS IN MANAGEMENT ACCOUNTING USING ARTIFICIAL INTELLIGENCE METHODS

Abstract

The article explores methodological approaches to organizing information flows in management accounting under conditions of digital transformation and the application of artificial intelligence (AI) methods. It substantiates the necessity of adapting the structure of synthetic and analytical accounting to the requirements of predictive and analytical models, as well as defining requirements for data sources, data quality, structure, historical consistency, and levels of granularity. A classification of management accounting data suitable for AI processing is proposed, including transactional data, master data and classifiers, planning and standard data, process and event data, costing and allocation data, external contextual factors, and unstructured or semi-structured data. The study develops a staged model of information flow movement from primary sources to the formation of KPIs, analytical services, and forecasting tools. Special attention is paid to the reconciliation of synthetic and analytical accounting levels as a prerequisite for the reproducibility of management calculations, the stability of performance indicators, and the correctness of machine learning (ML) datasets. The article also outlines the role of BI systems, particularly Power BI, in data preparation, quality control, semantic alignment of indicators, visualization of results, and integration of AI models into budgeting, costing, planning, and controlling processes. Considerable attention is given to control points within the data flow, versioning of rules and indicators, and the traceability of transformations required for the reliable use of AI models in business practice. In addition, the paper presents a practical perspective on the interaction between accountants, analysts, and digital technologies in the use of AI-based tools to support managerial decision-making, improve plan-fact analysis, identify anomalies, and enhance the interpretability of deviations in costs, revenues, and margins. The practical significance of the study lies in the development of an organizational framework for embedding predictive and analytical algorithms into the management cycle without losing the economic substance of accounting information, thereby improving the validity, transparency, and timeliness of managerial decisions.