Wrapping Up

Voice applications are an exciting next step in the evolution of human-computer interaction. Alexa, one of the leading voice platforms, enables users to use the natural interface of voice to perform tasks, ask questions, and be entertained.

Skills are applications that expand Alexa’s capabilities beyond her native functions. Skills can provide information, tackle business tasks, entertain, or anything you can dream of. The Alexa Skills Kit (ASK) provides tools, APIs, SDKs, and examples to help software developers produce voice-first applications.

Skills are composed of an interaction model, which describes how a human may interact with the skill, and a fulfillment implementation, which executes the logic to process questions and commands from the user. The interaction model itself describes utterances, samples of things a user may say to the skill, mapped to intents, which precisely capture what the user intended to say.

One of the most useful tools you’ll use as an Alexa skill developer is the ASK CLI, a command line tool that provides many useful functions for developing skills, including bootstrapping a new skill project and deploying the project.

Another useful tool that ASK makes available is the developer console where you can test a skill through a web interface rather than using an actual Alexa-enabled device. Even so, manually testing with the developer console is only one way to test skills.

Before we get too carried away learning new techniques and building new functionality into our Alexa skill, it will be helpful to get to know some of the options for testing Alexa skills. Therefore, in the next chapter we’ll look at several ways to test Alexa skills, both manual and automated, and even including ways of testing a skill without deploying it to the Alexa platform and AWS Lambda. This will give us the tools to consistently ensure that the code we’re developing does what we expect it to do.