best way to build ai chat with infinite memory with vector search ?
Hi, need some advice here.
i just want a simple infinite memory group chat, no file/image or pdf upload needed.
there'll be tables like
1. sessions
2. sessionUsers
3. messages
im thinking to create a vectorIndex on the message table.
looking at api, it have search the table by embedding which will return me list matching record (id & score)
right now I'm struggling to connect the dots here. How do I perform the vector search so that it would able to identify that which message belongs to who & when it is sent.
2 Replies
Thanks for posting in <#1088161997662724167>.
Reminder: If you have a Convex Pro account, use the Convex Dashboard to file support tickets.
- Provide context: What are you trying to achieve, what is the end-user interaction, what are you seeing? (full error message, command output, etc.)
- Use search.convex.dev to search Docs, Stack, and Discord all at once.
- Additionally, you can post your questions in the Convex Community's <#1228095053885476985> channel to receive a response from AI.
- Avoid tagging staff unless specifically instructed.
Thank you!
https://docs.convex.dev/search/vector-search#defining-vector-indexes Define filter fields with the index, eg.
chatId, sessionId, userId etc, depending on how you organise your data. Then you can restrict the data returned based on whatever grouping your want to search for.Vector Search | Convex Developer Hub
Run vector search queries on embeddings