The integration of Ukrainian enterprises into the European economic space, the attraction of foreign investments, and the reform of accounting and management policies require domestic business entities to ensure a high level of transparency and control. However, a significant number of Ukrainian enterprises still rely on outdated internal control tools. These limitations hinder the detection of hidden risks, the evaluation of management efficiency, and the timely formulation of managerial decisions. This article presents the results of a study aimed at detailing the architecture of an intelligent internal audit support system (IIASS) for business entities. Based on an analysis of the system's functional requirements, three levels of architectural modeling are proposed: a high-level component structure, a detailed transactional data processing scheme, and an integration schema with corporate systems. The proposed computational core structure includes modules for preprocessing, normalization, and feature generation, as well as the implementation of ensemble forecasting models based on logistic regression and XGBoost. Special attention is given to input data validation and the automatic retraining of models when significant deviations in feature structure are detected. The developed API integration mechanism with external systems (CRM, ERP, document management) is based on REST architecture principles and includes authentication methods such as OAuth 2.0. The proposed architecture ensures flexibility, scalability, and interpretability of decision support processes in auditing. The obtained results may serve as a methodological foundation for the practical implementation of intelligent audit systems in corporate environments.
Hnatchenko et al. (Thu,) studied this question.