A proof-of-concept project using RL to learn an agent policy of fire dynamics using reinforcement learning.