Text this: Optimization of pour-over coffee extraction variables using reinforcement learning