0

How can I program the VXML application so that I can recognize everything the user says and then what he says to send to another system?

At this point, I'm trying to use the tags but it doesn't give the desired effect. Maybe I do it wrong, or there is another way.

Michal Robak
  • 19
  • 6
  • Please provide your use case and what you have attempted. From the current content, the question isn't clear. If you are trying to support a dictation solution, you should also identify the platform (IVR and speech engine) used as you would be looking at proprietary extensions as the standards don't support that type of input. – Jim Rush Nov 18 '19 at 14:36
  • Hi, thank for reply I currently have a Dictation engine connected via MRCP. I wrote a script in VXML based on tags that reads what someone said, but only when the word is in grammar. My case is different. I want to send every transcription that the Dictation engine returns to me. It doesn't matter if it's in grammar or not. Unless there is another way to use the Dictation engine connected via MRCP in VXML? – Michal Robak Nov 19 '19 at 19:57
  • Given the nature of your solution, it is more about the integration than the standards. The standards only account for a structured grammar returning matches. Platforms, when they provide features that are in the standards often have work arounds. For example, a dictation engine, could always match and just return what it has. But, you are going to need to get this info from the vendor(s). – Jim Rush Nov 20 '19 at 19:57
  • I have already connected Dictation Engine with Asterisk (actually Voximal) The main question is how can i use Dictation (via MRCP) during call for every user input. – Michal Robak Nov 21 '19 at 17:58
  • Voximal use 2 different connectors for Dictation / Transcription. One embedded a specific plugin module to support GRPC / and other streaming protocols. Another is using the MRCP connector, generally in this case you need to use a builtin grammar. – Borja SIXTO May 11 '20 at 23:48

0 Answers0