Scalable cooperative multiagent reinforcement learning in the context of an organization