Chapter 3. Getting Started with AVS

Now comes the fun part! The preceding chapters presented voice interfaces at a high level. We took a look at some design considerations and covered some of the currently available services and frameworks. Starting with this chapter, during the course of the remainder of the book we will get our hands dirty with hardware tinkering and programming some code as we work on getting our voice interface prototype up and running.

We will begin with creating our very own Alexa-enabled device using a Raspberry Pi 3, a microphone, and a speaker. As we mentioned in Chapter 2, Alexa Voice Service (AVS) enables us to leverage Amazon’s intelligent voice platform on our own custom hardware.  Essentially,  we will be building our own version of an Echo or Dot, just not quite so fancy. There are some limitations with AVS compared to an off-the-shelf Amazon device, such as not being able to make calls or use the device as an intercom. In addition, when a new feature is released on the Alexa platform, it is not always immediately supported on AVS. Nevertheless, AVS enables makers and companies alike to create uniquely branded voice-enabled devices, using the power of Amazon’s voice AI. When we’re finished, you will be able to ask Alexa for the time or weather, request a song to be played, or even ask her to tell you a joke. A full list of AVS-available features can be found at

In subsequent chapters, after we have experimented with ...

Get Programming Voice Interfaces now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.