Issues of Developing an Intelligent System for Preventive Monitoring of the Occurrence of Operational Risk Events

As part of operational risk management, business processes are considered as one of the generators of risk events. Empirical data show that a significant share of operational losses in credit institutions is associated with the human factor, namely, the professional unpreparedness of employees. In this context, a promising area of risk management is the development and implementation of tools based on artificial intelligence, designed for automated assessment of the criticality level of operations, which allows mitigating the risks generated by personnel. The aim of the paper is to develop an intelligent system for preventive monitoring of the occurrence of a critical state of a business process due to actions or omissions of personnel to prevent an operational risk event. To achieve this goal, professional and personal criteria for evaluating personnel, criteria for assessing their impact on the business process, as well as accumulated statistical indexes were analyzed. The general structure of the business process status indication system is proposed, organized according to the modular principle. It is proposed to use artificial neural networks (ANN) of direct propagation as composite modules. The paper describes the main data flows coming to the ANN inputs and compares different ANN models for each of the system modules. The results obtained can be used in various areas of activity related to personnel actions to prevent the negative consequences of the critical state of the business process.