Programming Voice Interfaces

Chapter 4. Iterate: Evolve the Prototype

So now that we have a working voice-enabled piece of hardware for us to work with, we’ll want to start evolving it and customizing it to fit our use cases. First, let’s get our feet wet and start with a basic “Hello, World” Amazon Alexa skill. Once we have that in place, you will have a better understanding of how to create skills, how to use intents and slots, as well as how to handle sessions, exceptions, and the various request types (e.g., LaunchRequest, IntentRequest, and SessionEndedRequest).

Why Iterate?

The evolution of a product isn’t a single cycle occurrence. This is something that requires constant attention and fine-tuning. Development cycles should be kept short and nimble. This allows us to rule out issues fast, which is where the common mantra of “fail fast” comes from. Issues can arise from various points along the product lifecycle, from technical issues to user experience.

At this point, you’re simply experimenting with the prototype we’re building in these pages and learning as you go. But in the long run, you will want to plan for maintenance even if you’re just building a device for yourself rather than something for market. We will touch on basic over-the-air firmware updates to cloud-based services updates in Chapter 6. This will help you set up efficient methods for updating your voice-enabled IoT device long into the future.

Intents, Utterances, Slots, and Invocation Names

Before we dive into creating our ...