Yeah! I couldn't have described it better than you just did!

Feel free to take a look at the code and star the code if you like it. I'll keep upgrading it and adding more Reinforcement Learning variations