We build a cache for cache for ChatGPT, see how it can save your money and time when build application on top of LLMs