Skip to main content

Telephone channel

caution
To create a phone channel in a project you need to have added at least one connection.

Connection

Navigate to the project and click Channels in the toolbar. In the Telephony section, click Connect channel.

Fill in the fields:

  • Name — specify the channel name.

  • SIP trunk connection — select a connection. The connection is created during the telephony setup.

  • Allow inbound calls — enable the option if you want inbound calls to be received.

    caution
    You can create multiple outbound channels for a single connection. There can be only one channel that will accept inbound calls per connection.
  • Record calls — enable the option if you want to record and store the calls made. A call audio recording is stored for 6 months. The log record for a call is stored for a year.

    caution

    A call audio recording can be downloaded from the URL, which does not require authorization in the platform.

    If JAICP is installed in your infrastructure, we recommend monitoring requests to any URLs of the https://{host}/api/recordings/token/… format or limiting access to them.

  • Branch — specify the project branch you want to deploy into the channel (master by default). You can also select a Git tag or enter a specific commit hash.

  • Deployment — select Automatic (whenever any change made to the project is saved) or Manual (using the Deploy button in the channel description line).

    caution
    You have to deploy a script before starting the bot. At this stage, the system builds the bot, checks the script syntax, and performs tests.

Go to ASR and TTS, or press Create. In this case, the default speech recognition and synthesis providers will be selected, based on the project NLU language:

  • Projects in Russian will use Tinkoff ASR and Yandex TTS.
  • Projects in all other languages will use Google for both ASR and TTS.

ASR and TTS provider selection

Bots that make and receive calls use speech synthesis and recognition:

  • Text-To-Speech (TTS) or speech synthesis is the process of speech generation from printed text.
  • Automatic Speech Recognition (ASR) or speech recognition is the process of converting speech into text.

When creating a telephone channel, you can select providers for ASR and TTS. To do this, go to the ASR tab and select the connection, then do the same for TTS.

Please note that if you choose a specific provider for ASR and TTS, in case of problems, you need to switch the channel yourself to another provider.

You can also leave the option By default, in this case the settings of the most stable ASR and TTS providers will be automatically applied. In case of problems with providers, the channel will be switched to another provider.

Press Create. The created channel will be displayed in the list.

Editing

To edit the data of a telephone channel, click on the  icon, select Edit channel. Make changes and click Save.

Deleting

To delete a channel, click on the  icon, select Remove channel, then confirm deletion.