This article proposes the development of a multi-agent system with cognitive properties designed to find optimal solutions in the work environment. The Multi-Agent system consists of two agent sub-agents that contribute to maximum profits for their team. Aggregation subsystems are devoid of functions the communication capabilities. The status of an agent can only be identified in terms of his actions on the business environment. The process of finding an optimal solution is to solve the Nash equilibrium condition. The operating mode of the system is based on the structure diagram, the cognitive agent model and the knowledge and knowledge building diagram. The implementation is intended to use MCUs. Evaluation of the Nash equilibrium conversion rate was performed for 4 cases: one agent in each subunit, 5 agents in each subunit, 10 to 5 and 5 to 10 agents in each subset. The results are presented in the form of convergence graphs.
