ell is a lightweight prompt engineering library treating prompts as functions, that enables prompt versioning, optimization, tracing, readability, and visualization via lexical closures.

I built ell based on some ideas during my time as a research scientist at OpenAI around language model programming, with the aim of building the PyTorch of prompt engineering. AI engineering needs good, open-source and free tooling, so we've built a tensorboard-like visualization tool (studio) packaged along side ell to fully leverage this new library.

really excited about this, and would love some feedback!