Wrapping Up

Although Alexa has a lovely and very natural voice, we can apply Speech Synthesis Markup Language (SSML) in our skill’s responses to alter the way she speaks, pronounces words, and even change her voice to a different voice altogether.

Using the text-to-speech simulator in the Alexa developer console, we can try out snippets of SSML and hear how they will sound before adding them to our skill’s fulfillment code. Once it’s ready, we can inject the SSML as text to be returned in the response or as prompts in the interaction model. Applying this to the Star Port 75 Travel skill, we were able to add sound effects and music in the response to the launch request as well as spice up validation prompts with a few interjections (known as “speechcons” in Alexa terminology).

While SSML is very powerful, it can also be quite verbose. Speech Markdown offers a more terse approach, enabling you to write rich speech responses using a variant of the popular Markdown text processor to produce SSML responses.

SSML can do amazing things to add emotion, personality, sound effects, and customize how Alexa speaks. In the next chapter, we will take the sounds and speech produced by SSML and returned from Alexa skills to the next level by applying the Alexa Presentation Language for Audio to combine speech, sound, and music into layered and rich audio responses.