A smooth voice experience requires several different pieces of technology working together. The three primary systems involved are the speech pipeline, responsible for listening to the user and converting their speech into text; Natural Language Understanding, which translates phrases and sentences into actionable input for an application; and Text to Speech, which converts textual system responses into audio to be read to the user.

These pages exist to introduce, in a platform-indepedent way, how Spokestack uses these technologies and lets you configure them. They also cover general topics like converting an NLU model from other voice assistant platforms into Spokestack’s format and even how to design custom wake word models for use with Spokestack.

If you’re looking for code samples, please see the section dedicated to your mobile platform of choice.