I am working on hugind, I have two goals:

- make it reliable to run LLM inference on company hardware, even when it is poor or outdated

- bring chaotic agentic behavior under control in business contexts

https://github.com/netdur/hugind