This page demonstrates a call to speak() that will be deferred and will be executed instantly after the configuration and any voice definition has been set up.