Hi everyone! We are going to be live in closed beta pretty soon for my new project NeroBot, a next generation web crawler to support Enterprise RAG apps. We have been very inspired, from early on, by the LlamaIndex framework and research, thank you team. (hopefully this is the write channel)

Check out our website and get on the waitlist here: https://www.nero.ai/

Also i wrote a brief post outlining the story for why we are building: https://blog.nero.ai/why-were-building-nerobot-the-next-generation-crawler-784e5150f3bb

Has anyone tried getting llm to write the pydantic model?

I have everything working with pydantic and extracting desired fields, but now i want to enable the end user to provide natural language for the desired extraction and then have step where the llm defines the pydantic model

Would be great if it could be just an easy json to class function etc