The RL algorithm is complete, thoroughly tested on an OpenAI environment (record some videos for demos or future slides!). At this point, we should have an in-depth understanding of an RL algorithm along with a general idea of its pros and cons.