This paper examines the hybrid architectures of decision-support systems in organizations, including a decision maker and a neural network artificial intelligence system. The paper examines architectures in which the neural network of an artificial intelligence system is trained by the experience of a decision maker, as well as artificial intelligence systems trained by the states of a subject area to identify patterns of its development and support decision making in the field of business analysts. The architecture of a decision-support system aimed at stimulating the intuition of the decision maker when working with unstructured data is considered. The proposed criteria system provides a qualitative analysis of the considered architectures, their advantages, and the barriers to use in decision-making systems.