AI Chat Memory + App Data Question:
————————————————-
I see there are a few options for a plug+play chat memory node that hooks into the LangChain AI node.
Some of the options include
Redis
Postgres
and more
I’m trying to square how this chat memory would be associated with a sales lead that I’m calling via an phone AI service, and putting that phone call transcript in a Pocketbase database along with all the rest of my application’s data.
Maybe they aren’t related and this AI chat memory is actually just for me and department of AI agents to talk to each other “internally”?
What about emailing/texting/calling back the sales lead then?
I suppose what I’m asking is, is this memory store ONLY for recalling your own chat logs (aka internal business comms and AI prompts)?
Does the chat memory node really need to have a database all its own? Does that database need to be No/SQL or can I push via API endpoints for Pocketbase?
I suppose the highest level question is: can I just have everything in one database? I don’t want to break up anything.
Update: based on my research into LangChain docs for each Chat Memory module, I would have to use one of the Backends they suggest unless I want to create the connection code and the under the hood code for Pocketbase. Unless I’m wrong?
Update to: this just dropped right after I posted this perhaps this will be the standard?
Ensure to copy your n8n workflow and paste it in the code block, that is in between the pairs of triple backticks, which also could be achieved by clicking </> (preformatted text) in the editor and pasting in your workflow.
```
<your workflow>
```
That implies to any JSON output you would like to share with us.
Make sure that you have removed any sensitive information from your workflow and include dummy or pinned data with it!
Chat memory is used just for that, “to persist chat history in your workflow.” This will allow asking questions related to the previous question without repeating the question itself.
If you do not need the history of the previous conversations you simply do not add the memory sub-node to AI Agent node.
The issue is I do need it in one context, (talking to customers, industry knowledge), but not in the other (querying application data (and optionally, industry knowledge)).