Hi everyone! Sorry to bother you, I have a noob question.

In the company I’m working on we’re looking to implement a LLM chatbot that could answer to our client questions using pdf documents.

But also we want this chatbot to trigger some API calls in order to update some rows in our database using OpenAI function calling.

This frameworks looks very promising but I want to ask you guys if you think the state of art of LLM chatbots are ready for production (external usage, like customer support), I’m very scared of the hallucination thing or our chatbot inserting fake info in our database.

Can someone please give me its honest opinion about this?