One Minute Research in Plain Language: Making reinforcement learning efficient in real-world systems