One of the most important breakthroughs in AI research is deep reinforcement learning (RL), which underpins superhuman play in complex games such as chess and Go. At Arena, we use deep RL to optimize and automate business processes.