I made a web app for chatting with all the content related to Confluence Space using Langchain.

[Pooja Soni]

Title: How to Automatically Generate Embedding for Updated Documents in a Confluence Space and Enable Real-Time Question Answering on the Updated Data?

Body: I'm currently working on accessing a Confluence space using Lang chain and performing question answering on its data. The embedding of this data are stored in a Chromadb vector database once I provide user name,API key and Space key.
However, I'm looking for a way to automatically generate embedding for any documents that change in real-time within the Confluence space and enable real-time question answering on the updated data. Any suggestions or solutions on how to achieve this would be greatly appreciated!

[Rafal Wasowski]

Hi Pooja,

I just came across this thread whilst searching the web to find out if anyone has successfully attempted this. Have you made any progress since you posted this?

I don't have admin privileges so I don't think I can create webhooks, but that sounds like a good way to trigger a (re-)creation of a vector embedding when either new content is posted or old one is updated.

In the absence of webhooks, I'm toying with the idea of utilising the RSS feed as a compromise. It does have all the relevant data (URL link, creation time etc.).

It would be interesting to learn where you got to with your project if you were happy to share.

[Nic Brough -Adaptavist-]

Welcome to the Atlassian Community!

Why are you doing this?  Confluence is already a wiki that provides you with real-time access to all of its updated data, and if used as a knowledge base, suggests pages that answers questions to your users.

Could you explain what problem you are trying to solve here?