The article is devoted to the problem of improving the efficiency of managerial decision-making in the agro-industrial complex under the conditions of digital transformation. The current stage of agricultural development is characterized by the intensive implementation of digital technologies, which leads to a sharp increase in the volume and diversity of data available for analysis and forecasting. Under these circumstances, traditional management methods based on retrospective assessments and expert experience demonstrate limited effectiveness and require a fundamental revision. The article substantiates the need to move from static and linear models to adaptive, self-learning management systems that integrate analytical, cognitive, and predictive components. Particular attention is paid to the role of artificial intelligence in ensuring the stability and predictability of production business processes, as well as to maintaining the importance of the human factor in data interpretation and strategic decision-making. The purpose of the study is to develop and substantiate a hybrid algorithm for managerial decision-making that integrates the practical experience of agricultural specialists with the computational capabilities of artificial intelligence. The algorithm is based on the principles of adaptability, cognitive compatibility, and iterative learning, which makes it possible to increase the accuracy, predictability, and resilience of management processes under conditions of high uncertainty in agricultural production. The developed algorithm was tested in a model experiment using production monitoring data, which confirmed its effectiveness and applicability in the practice of digital management of agri-food systems.