The community member is asking how to turn a Q&A RAG (Retrieval Augmented Generation) into a multi-turn chat experience, where the initial question retrieves context and subsequent questions continue the conversation like a chatbot, using the template and context as the system message.
In the comments, another community member suggests trying the ChatEngine, providing links to relevant documentation. Another community member asks if the ChatEngine's memory is multi-user safe, as they have this running behind an API. A third community member mentions planning to persist the chat history in a PSQL JSONB database. Finally, a fourth community member states that they have managed to get something working, but notes that the cache is not for production use.
There is no explicitly marked answer in the provided information.
