How can we query the pre-made Chatgpt Assistant's thread instead of creating new one for each workflow?

Describe the problem/error/question

We want to create an automatic content generation for one brand, and we want those regular content post should be consistency based on our content direction.

What is the error message (if any)?

Please share your workflow

Share the output returned by the last node

We want the out put is the content of the AI after it read our documents and requirements.

Information on your n8n setup

  • n8n version: 1.22.4
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main): don’t know
  • Running n8n via (Docker, npm, n8n cloud, desktop app): destop app
  • Operating system: windows 11

We want the out put is the content of the AI after it read our documents and requirements.

Hi @mivtool, welcome to the community!

You’re probably looking at a Retrieval Q&A Chain here: Retrieval Q&A Chain | n8n Docs

This allows you to answer questions against a number of different source, the most common one being a vector store. But you could even use an n8n workflow providing data as your data retriever.

Imagine you have a workflow querying your knowledge database and returning a result like this:

You could build a simple Question and Answer Chain like so:

This is the result:

But this really is a very simple example. I suggest reading the documentation around this topic, depending on the amount of data you’re processing a vector store is probably a much better fit (but harder to visualize, hence the simplified example above :wink:).


Thanks for you share.
I saw some langchain workflow like that and follow but It’s lil’ complicated to me, seem the workflow now is different from the last 2 month - the time that the clip was published.
Hope that this can help easier.

1 Like

Hi @mivtool, I agree - this is complicated for me too. LLM technology (and AI in general) is developing at an incredible rate, so sometimes it’s hard to keep up to date with everything happening. I hope n8n makes this at least a little bit easier for you, and never hesitate to ask if something comes up!

1 Like

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.