Hi folks, I recently set about to optimize Hamilton's memory overhead -- shared my learnings and I think it has some interesting lessons that python developers can learn from. Feel free to ask any questions (or point out anything I might have missed!).