Intelligent systems for planning and optimizing enterprise business processes

Abstract

The paper investigates intelligent planning and optimization systems for enterprise business processes as an element of modern process management. Business process intellectualization is interpreted as a change in the management cycle logic: from regulation and execution control toward data, deviation diagnostics, decision outcome forecasting, and computationally grounded selection of optimal actions under resource, time, and quality constraints. The conceptual apparatus is systematized: enterprise business process, process approach, process planning and optimization are differentiated; distinctions between automation systems, expert systems, and intelligent systems are established; differences between BPM, ERP, and APS systems are clarified. The evolution of the BPM paradigm – from workflow systems and RPA to iBPMS and hyperautomation – is examined. Three intelligence contours are substantiated: deviation diagnostics (process mining), risk prediction (predictive monitoring), and prescriptive intervention recommendations. Intelligent planning methods are analyzed: operations research, machine learning, reinforcement learning, and the role of process mining as a diagnostic bridge between BPM and optimization. Organizational and economic implementation effects are identified alongside key success conditions – data governance, process ownership, and a continuous improvement culture. The sociotechnical nature of AI/BPM project failures is substantiated. The specifics of the Ukrainian context are investigated: resilience under wartime conditions as the primary digitalization driver, SME investment constraints, cloud-modular architecture, and EU integration requirements under the AI Act and Data Act. The authors' position: the key strategy is building a managed process-data foundation on which AI and OR methods become tools for resilience and EU integration compliance. Practical readiness criterion: if a process is unmeasured and has no owner, it cannot be optimized – only its chaos can be automated.