This paper introduces a novel method to improve the decision-making process of reinforcement learning agents via quantum information technology methods. In this approach, states |s〉 of the system are replaced by quantum states (eigenfunctions) of the system |ψ〉, in which a system can be in a superposition of states, and rewards of each step are calculated based on the calculated eigenvalues of the previous step. The agent decides the next step of the system based on the result of the quantum gate’s effect on the available options. Due to the high-speed, escalated performance of quantum algorithms, this method will improve the performance of reinforcement learning agents in unknown environments.