This post discusses techniques for using large language models (LLMs) as a mathematics assistant to derive equations modeling the behavior of software processes. The author shares personal experiences and insights applicable to both software and non-software contexts.