THE EMPLOYEE TRAINING BUSINESS PROCESS IN AN IT COMPANY: DIAGNOSING THE CURRENT MODEL AND LEVELS OF MANAGERIAL DECISION DELEGATION TO ARTIFICIAL INTELLIGENCE

The article examines the employee training business process of an IT company as an object of process management under the growing influence of artificial intelligence (AI). The purpose of the study is to formalise and diagnose the current model of the training process, to establish its systemic limitations, and to develop a classification of AI functions by the level of delegated managerial decision. The methodological basis comprises the process approach, systems analysis, functional decomposition, lean waste analysis and expert estimation. A process passport is compiled in a diagnostic format, an AS-IS flowchart is constructed as a linear sequence of nine stages, and a quantitative decomposition is performed: the full cycle lasts 9–12 months, while the share of stages that add value directly to the employee does not exceed 15–20 %. Waiting and unused employee potential are shown to be the dominant categories of waste, and the open-loop character of the control circuit is identified as the key systemic limitation. Four root causes of inefficiency are established, and it is demonstrated that they form a closed causal loop reproducing itself irrespective of the quality of learning content. The scientific novelty consists in the formalisation of the training process passport through the category of the competence gap, and in a classification of artificial intelligence functions by the level of delegated managerial decision (L0–L4) whose criterion is not the technical sophistication of the model but the scope of authority transferred to the algorithm; a normative rule is justified prohibiting the delegation of decisions with personnel consequences above level L2. Practical value: the process passport, the waste map, the list of root causes and the delegation scale constitute a diagnostic toolkit suitable for auditing the learning and development function and for designing an artificial intelligence usage policy in human resource management. It is advisable to direct further research toward verifying the current-state model and designing an improved process model capable of eliminating the identified cause-and-effect loop.