yeah i want to use llama index to store the index and use langchain to read the index and then build a chatbot with memory history, so i cant use index.as_query_engine()( i think it's doenst remeber the previous conversation?), that 's not easily possible?