Intellectualization of the enterprise administration system: from information support to predictive-prescriptive management

Abstract

The purpose of the article is to substantiate a methodological approach to the intellectualization of enterprise administration through the economics of managerial decision-making, rather than by the number of AI tools implemented. The methodology combines systems and institutional perspectives, decision theory under risk, stochastic optimization, and scenario analysis. An administrative decision episode is proposed as the basic unit of analysis, linking an information signal, uncertainty scenarios, feasible alternatives, resource and regulatory constraints, decision rights, an economic loss function, and the realized outcome. The net economic value of intellectualization is defined as the difference between expected losses under baseline and intellectualized decision procedures after deducting the costs of AI infrastructure, governance, oversight, and organizational adaptation. The decision rule accounts for expected losses, conditional value-at-risk, implementation costs, and the economic cost of time. Administratively feasible actions are constrained ex ante by budget, authority, compliance, data quality, explainability, and acceptable risk. Human participation is differentiated by decision class according to the scale of consequences, reversibility, uncertainty, and sensitivity. For organizational learning, total decision loss is decomposed into components: information, forecasting, choice, approval, and execution. Scientific novelty lies in shifting assessment from technological maturity to the measurable economic value, admissibility, and accountability of a specific managerial decision. Practical significance consists in defining boundaries of algorithmic delegation and identifying sources of unrealized effects. For construction-development enterprises, the framework supports financial, resource, scheduling, procurement, and multiproject decisions in which analytics should change not only forecasts but feasible actions. Further research should empirically calibrate loss functions, risk thresholds, and comparative effects of conventional and intellectualized procedures. Keywords: enterprise administration; managerial decision; predictive analytics; prescriptive analytics; decision loss; algorithmic delegation; digital transformation.