We are using Google speech to text API from avaya OD. It works fine. We make use of record tag to record user speech input send the audio file using Google speech API to convert it into text .HOWEVER, the recording of user input starts after playing the initial message. For ex : You can say things like billing , payment.....please say the reason for your call. After this message the recording starts.
The problem or question is , how do we start collecting users input , if the caller starts saying billing once the prompt starts. Ex you can say things like billing... And user says billing. The recording starts only after the prompt completes. We cannot play the message in prompt and collect node , coz we need to send the audio file to Google to transcribe to text.
How can we achieve this without an ASR. Also does google ASR works on prem ?
Unfortunately we cannot barge in from record tag. Instead use prompt_and_collect and with recordUtterance property to true. This can be set in input form items. You will get the utterance of the speech the user said, it allows bargein. Also remember to set the grammar file to "provider": "google"