Conceptual model of building a supply chain management system based on artificial intelligence tools

Abstract

This article presents a conceptual model for developing a supply chain management system (SCMS) based on artificial intelligence tools. The relevance of this research lies in the fact that artificial intelligence is often used in supply chains for specific tasks, such as demand forecasting, route optimization, or improving warehouse operations, while the principles for integrating these into a single end-to-end system at the corporate level are insufficiently systematized.

The objective of this study is to substantiate the architecture of an AI-based SCMS that integrates planning, execution, and management modules, includes a data layer, an analytical (modeling) layer, a decision-making cycle, and a feedback mechanism, and to propose a system of key performance indicators that simultaneously measure model quality and business impact for performance evaluation.

The methodological framework utilizes methods of system analysis, literature review and synthesis, conceptual modeling, and comparative logical analysis. Results - the study structured the SCMS architecture at the "data-model-decision-feedback" levels, describing the main types of models in the analytical layer (forecasting, optimization, risk and anomaly detection, scenario modeling) and their lifecycle in a production environment.

Within the decision support mechanism, human-system interaction modes were defined: consultation, collaborative decision-making, and independent execution. Furthermore, a system of key performance indicators was proposed in two dimensions: assessing model quality and measuring its impact on business results. Conclusions - the proposed model provides organizations with a methodological basis for the phased implementation of artificial intelligence in supply chain management systems and a substantiated assessment of its effectiveness.