User interface is primarily spoken voice. An initial plan: Prototype in Python with LangChain Port to a mobile app?