I just asked the same question Running 1

tbh I think it would be cool to make a llama_index_server repo or similar. Lower priority but I have it on my list

Yeah my app is basically ready to go except this piece is missing. I really don’t want to rebuild the context and pull the chat history every time I get a new message. I’d like the ability to create llama index user sessions that persist for some time and then get torn down as needed

any chance you know of an example that demonstrates this?

The create-llama app is one example of this (examples with both python and JS)

Thanks will take a look under the hood