Conceptual Model of a Multi-agent System to Support the Activities of a Tender Specialist

The paper considers the task of constructing a conceptual model of a digital twin of a tender specialist based on an intelligent multi-agent system focused on the search and evaluation of the relevance of procurement by the supplier. The relevance of the study is due to the growing volume of procurement information, the complexity of the structure of tender documentation and the high cost of errors at the stage of pre-selection procedures. It is shown that in the supplier's practice, procurement search is not a technical keyword filtering operation, but a multi-criteria analytical process that includes interpreting the search task, analyzing documentation, checking participation restrictions, evaluating the customer, taking into account the terms of the contract and comparing the purchase with the company profile. Based on the analysis of the professional activity of the tenderer, functional and conceptual representations of the contour in question have been developed, allowing it to be described in the "as is" and "as it should be" states. It is proved that the most structured part of the professional circuit can be transferred to a digital format in the form of an intelligent multi-agent system. A conceptual model of the digital twin is proposed, which includes an analytical core and supporting modules for management, profiling, setting evaluation policies, retrospective analysis, validation of results, interpretation and formation of the final presentation of the purchase. The results obtained can be used as a basis for the software implementation of the digital twin and its testing in a professional environment.