Ytel, a communications software company, has integrated its communications platform-as-a-service (CPaaS) with Google Cloud Platform to take advantage of Google's Cloud Text-to-Speech and Speech Synthesis Markup Language (SSML).
Ytel is developing new capabilities using the Google Cloud Platform for interactive voice-response (IVR) systems and automated help assistants that incorporate DeepMind WaveNet voices that are near human quality.
Ytel is one of the first CPaaS companies to adopt Google Cloud Text-to-Speech. TTS, DeepMind WaveNet Voices, and SSML enable developers to create interactive voice applications that dynamically generate speech rather than playing static, pre-recorded audio files.
"We are thrilled to be able to bring WaveNet and other synthetic voices of the Google Cloud Platform into our CPaaS. This will dramatically improve the quality of the voices offered to our clients to meet their personal or client engagement goals," said Matt Grofsky, chief technology officer at Ytel, in a statement. "We envision many new applications that can leverage the feature set of the Google AI and Machine Learning Platform. For example, consumers can expect to see new apps that will be able to take more advanced actions within CPaaS, such as machine learning-based workflows to drive conversations."
The Ytel communications solutions now support the more than 170 Google Cloud Text-to-Speech voices in more than 60 languages and dialects. More than 80 of these voices are WaveNet-generated, which according to Google, "mimics stress and intonation in speech by identifying tonal patterns" and "produces much more convincing voice snippets than previous speech generation models." Ytel's use of Google Text-to-Speech and SSML also brings more control over the synthesized speech output, including volume, pitch, rate, and pronunciation.