How would you architect the infrastructure to support a new generative AI chatbot for a client's customer service platform?
