description: For the Energy Transition to be timely and feasible, functioning electricity markets are imperative. Yet, existing modeling approaches could be insufficient to appropriately assess future electricity market design. In this context, Reinforcement Learning appears as an alternative to further improve the existing modeling framework, by allowing policymakers to assess more intricate elements in the agent’s decision-making process. This presentation will explore some of the challenges, obstacles and opportunities for Multi-Agent Reinforcement Learning approaches with the Ray Libraries.