Partial transcription

[jbl]

Hi,

I have tried to reproduce the video for partial transcription using the GitHub blueprint, I've also tried the python SDK. With the python SDK i see conversation events pushed over the websocket, but receive no transcribed results. Looking at in browser i see the active-conversations in the new widget, but no transcribed results(screenshot below).

The set up procedure in the videos and blogs have been followed, but one area I keep seeing is that in the Program tabs in the Genesys documents, I see they have transcription options but I see no "select transcription engine" or similar offered, as it shows in some of the images.

I have enabled transcription on the Queue, but this also says to be enabled globally as an Organization, which I can't find to confirm or deny.

Partial transcription inside inside an iframe script (just to make it appear center screen)

From what i can tell, the service is subscribing to the correct notification events? (copied from networks tab in browser)

{
  "entities": [
    {
      "id": "v2.conversations.xxxxxx-61d8d5a57a3b.transcription"
    },
    {
      "id": "v2.routing.queues.xxxxx-c7b01d79638.conversations.calls"
    },
    {
      "id": "v2.routing.queues.xxxxxx-cbddd0362fc8.conversations.calls"
    },
    {
      "id": "v2.routing.queues.xxxxxx-d639f9143fa5.conversations.calls"
    },
    {
      "id": "v2.routing.queues.xxxxxx-b9ff-05e870238b54.conversations.calls"
    },
    {
      "id": "v2.routing.queues.xxxxxx-8a28-8931f9f379e4.conversations.calls"
    },
    {
      "id": "v2.routing.queues.xxxxx-daf56e4cf484.conversations.calls"
    }
  ]
}

[Jerome.Saint-Marc]

Hello,

In order to leverage Transcription Notifications, you need to configure voice transcription which includes enabling Line Recording (Dual Channel) on the Trunk that you are using to get inbound calls.
With Transcription/Recording, the call must go through your Genesys Cloud Voice or BYOC SIP trunks - it must be initiated or received by an external caller/phone.

The topic for Transcription notification is: v2.conversations.{id}.transcription

Regards,

[jbl]

Thanks for your respond,

I'm still confused, I've enabled Voice Transcription on the queue that call is transferd to from architect. I'm certain that I've done every step in configure-voice-transcription. To me starting to feel like we are on version that does not support this? Or something with our license?

Ive gone over the conversation-ids, and to the best of my knowledge we are adding subscriptions to the right conversation ids as well. The python sdk gives me no errors, so i feel like code is right. There is just no text getting emitted on the ws-channels.

[Jerome.Saint-Marc]

What about your Trunk (Genesys Cloud configuration side - Admin - Telephony - Trunks)?
Are you sure that Line Recording and Dual Channel are enabled on the trunk that you are using to get inbound calls to Genesys Cloud? I am referring to the trunk configuration in Genesys Cloud Desktop/Admin.

Regards,

[jbl]

I do not think so, would that be the one called PureCloud Voice - AWS? I'm not to confident with how the trunks are connected/interact.

[Jerome.Saint-Marc]

It depends what you are using. I would recommend to ask your colleagues.
"PureCloud Voice - AWS" is indeed the name of the trunk if you are using Genesys Cloud Voice (i.e. a number purchased from Genesys).
You can also have other trunks if you are interconnecting with a 3rd party SIP Trunk provider (BYOC Carrier Trunk).
Enabling the recording on your trunk will not break anything (given you are in a lab env). So you can try to enable line recording and dual channel with one of the codecs/audio format listed in the link I included in my first reply.

Regards,

[jbl]

I've no idea why, or how. But that worked. I had to release one our DID-numbers and set that as outbound. But now, it works. Thanks for pushing me in the right direction.