AstriCon 2016

Earlier this year, Google made waves by announcing the availability of their speech recognition API as a paid service. This means that developers using Google’s API, for the first time, can build real applications on their platform without fear of the service being yanked or altered unexpectedly. Besides having excellent accuracy, Google also covers many languages. This is just one example - the entire speech recognition and “conversational UI” space is heating up. This talk is a review of the options available to developers and integrators. We’ll explore what works best for telephony and Asterisk vs. mobile apps and WebRTC. We’ll look at features and limitations of the various engines in each context. We’ll also demo this technology in action, in Asterisk, in WebRTC, and in an appliance: Amazon’s Echo.