AT&T has announced the launch of a new Speech API which is powered by the AT&T Watson℠ speech engine. The new Speech API allows developers to integrate speech-to-text functionality in mobile applications. The service works with almost any mobile phone and "across all other U.S. wireless carrier services."
There are a variety of speech contexts that can be used to voice-enable applications which includes Web Search, Business Search, Voicemail-to-text, SMS and Generic. According to AT&T, the Generic speech context "allows developers to address over 90% of US customers with a single API."
Using the Speech Service, audio files are sent to AT&T via the mobile application. The transcribed text of what was said by the user is then returned to the application. The API supports XML & JSON request/response body formats.
According to AT&T, the Speech API provides these benefits:
- No need to build and maintain your own speech libraries
- It's a cross-carrier solution
- Simply plug and play: Connect to our Speech API, send us speech, and we send you text
AT&T also offers native and HTML5 SDKs as a jump-start.