Call Feature Support

Blueworx Voice Response uses Call Features to link external resources to other applications to generate text to speech, perform speech recognition or act as a virtual assistant. This page will briefly introduce each supported feature and what it does.

Text To Speech

Text To Speech resources take a text prompt in VoiceXML and turn it into audio that Blueworx Voice Response can play out to callers. These can be dedicated servers or a cloud service.

Currently supported TTS services are:

• MRCPv2 - any vendor that supports communicating with the Media Resource Control Protocol version 2 over the Session Initiation Protocol (sip/sips). Common examples of vendors would be Nuance or Lumenvox. See TTS MRCPv2 Configuration.
• Polly - Amazon's Text To Speech offering hosted on AWS. See TTS Polly Configuration.
• Google - Google's cloud Text To Speech service. See TTS Google Configuration.
• Watson - IBM's cloud Text to Speech service. See TTS Watson Configuration.
• Eleven Labs - Eleven Labs run an online service that supports TTS requests. See TTS Eleven Labs Configuration.

Google and Eleven Labs are supported in overriding their call feature settings via CCXML hints on a <dialogstart> - See here for more information.

​

Automatic Speech Recognition

ASR resources listen to an audio stream that Blueworx Voice Response supplies to them from a caller and analyse them to detect matches to predefined grammars. When matches are detected the information is supplied back to a VoiceXML application which can then use that information as required.

Currently supported ASR services are:

• MRCPv2 - any vendor that supports communicating with the Media Resource Control Protocol version 2 over the Session Initiation Protocol (sip/sips). Common examples of vendors would be Nuance or Lumenvox. See TTS MRCPv2 Configuration.

​

Speech To Text

Speech To Text or STT resources function similarly to ASR resources but instead of matching a limited set of grammars use natural language processing to turn any speech into text.

Currently supported STT services are:

• Google STT uses google's cloud services - see STT Google Configuration.
• Watson STT uses IBM's cloud - see STT Watson Configuration.

​

Virtual Assistants

Virtual Assistants use machine learning to construct models of an interaction and the paths that a user might make through a call, gathering information as required. Virtual Assistants require the use of Text To Speech and Speech To Text resources to handle the interactions, though they may do so internally to the call feature or using BVR defined call features.

Currently supported Virtual Assistants are:

• Watson - uses IBM Watson's services to handle a call. See VA Watson Configuration.
• Google Dialogflow ES - uses Google's cloud ES services (formerly just known as dialogflow) to handle a call. See VA Google ES Configuration.
• Google Dialogflow CX - uses Google's cloud CX services to handle a call. See VA Google CX Configuration.
• Nu Workforce - uses Nu Workforce's services to handle a call. See VA Nu Workforce Configuration.