Hello I am implementing llamaindex ContextChatEngine. I am looking for ways to get a structured output from my chat engine. Is there a way I can use output parser/pydantic program like LLMTextCompletion with ContextChatEngine? Any suggestions please.