Part 1 Framework for improving conversational AI
Have you ever had a bad experience with a chatbot? Perhaps with a voice system that always tells you “please listen carefully—our menu options have recently changed” or a chatbot that never understands your questions. Hopefully you’ve had great experiences with AI as well—AI that seems like it knows you and proactively identifies your needs. What separates good conversational AI from the bad?
Conversational AI is bigger than ever, as companies look to improve customer experience and their own bottom line through this technology. Generative AI has rekindled interest in the technology and made it easier than ever to add intelligence to a chatbot. Many of these chatbots look great in a prototype phase and then falter in production. This part of our book sets the stage for building and improving conversational AI.
Chapter 1 introduces conversational AI and generative AI technology, their pros and cons, and how they complement each other. It also lays out the types of pain points many AI solutions run into. Chapter 2 shows you how to build and evolve a chatbot, layering on incremental complexity and capability, ending with a hybrid of traditional and generative techniques. Chapter 3 demonstrates how to objectively evaluate your AI and develop plans to improve it.