Hey guys,
im thinking about a scenario, but im not quite sure if it works like that.
I have a lot of documents in pdf and want to fine tune my ChatGPT and create a vector database with it. So the idea would be:
When i prompt in ChatGPT it something like “Show me the best restaurant in Germany but dont list the one’s from my document XY , because i have already been there”.
So at the same time it searches the web(like normal) and has all my documents in the back to get from aswell. Basically - ChatGPT+my Data.
Might be a bit off topic, but i saw there is “Upload file to Vector base” in the ChatGPT Module.
Is anyone experienced in this (has easier methods)? 
The best scenario for me would be:
- I upload files into my Google Drive folder
- It gets uploaded via make scenario to my ChatGPT Account.
- When i open chatgpt.com and start chatting/prompting it always has access to my documents and answer my prompts based on the documents+web search
Greetings and thank you very much in advance!