Keeping Data Private But Connecting To LLM

API
1

I have seen different ideas about how to integrate my data with OpenAI, but keep my data proprietary.

What is the current way to approach building a chatbot where my data is not shared with the entire LLM?

Is there documentation around how todo this?

What tools (langchain, etc) should I use todo something like this?

After reading, I found hugging faces and the usage of it with openai.

Would the best and most efficient way to keep my data private, is to run my own LLC with hugging faces, and then have OpenAI reach back to it for my specific data?

Can Plugins use hugging faces?

Thanks

2

You can explore https://open-assistant.io/
However the quality of open source model as of today is not as good as GPT3.5 or GPT4.
Some of the better open source model also is not available for commercial use as well.
So you need to be cautious when using the selected models.