Voice

Bright Pattern Documentation

Generated: 11/28/2021 5:46 am
Content is available under license unless otherwise noted.

Voice Tutorials Overview

This section provides tutorials on anything related to voice services.

Tutorials



How to Configure Conversational IVR

In this article, you will learn how to configure conversational IVR for your inbound voice service. Because Bright Pattern’s conversational IVR is the front end to a bot, most of the configuration will be done on the bot side (i.e., through Google, GoVivace, or IBM in your workspace), outside of Bright Pattern. This procedure will guide you through the steps to take after you have set up STT and TTS engines through your service provider.

For more information, see the Contact Center Administrator Guide, section What Is Conversational IVR?

Prerequisites

If you have not already done so, please complete these steps before proceeding:

Procedure

Step 1: Set up a Speech-to-Text engine through your service provider

Step 2: Add Speech To Text Integration Account

A Speech To Text integration account uses your credentials to connect Bright Pattern Contact Center to a third-party STT engine. Integration accounts are configured in the Contact Center Administrator application.

  1. In Contact Center Administrator > Call Center Configuration > Integration Accounts, add a new Speech to Text integration account. For property descriptions, see the Contact Center Administrator Guide, section Speech To Text Integration (STT).

  2. Note that if you do not see Speech to Text as an option when adding a new integration account, that option has not been enabled for your contact center. See your service provider for more information.

Step 3: Set up a Text-to-Speech engine through your service provider

Step 4: Add Text To Speech Integration Account

Likewise, a Text to Speech integration account uses your credentials to connect Bright Pattern Contact Center to a third-party TTS engine. TTS integration lets you create voice prompts in numerous supported languages to automate bot responses.

  1. In Contact Center Administrator > Call Center Configuration > Integration Accounts, add a new Text to Speech integration account. For property descriptions, see the Contact Center Administrator Guide, section Text To Speech Integration (TTS).

    Select the type of TTS integration account


  2. See the Load Voices property. If this is a Google TTS integration account, load voices by completing an additional step in Google. See How to Load Voices from Google Cloud Text-to-Speech API.



    Google TTS integration account properties


  3. Then click Load Voices. This may take a few minutes. When the voices are loaded, a message will display the date they were added: “Loaded on XX/XX/XX.”

  4. Click Apply to save your changes.

Step 5: Add a Voice Scenario That Includes Play-Listen and Bot Scenario Blocks

The Play-Listen scenario block is a key part of conversational IVR. This block uses the integrated TTS and STT accounts that you just set up to play voice prompts to the caller and to listen for the caller’s response.

Chat Bot Select Account and Ask a Bot blocks allow you to choose which integrated bots will be used for interacting with callers.

  1. In Contact Center Administrator, go to Scenarios > Voice and import our conversational IVR scenario template: File:App Conversational IVR Example.zip.

  2. Work through our scenario-building exercise, How to Use Conversational IVR in a Scenario.

  3. Modify the scenario to suit your needs, being sure to select the correct bot and STT account in the scenario blocks.


Conversational IVR scenario example


Step 6: Add or Select a Scenario Entry

The scenario entry is the entry point for the caller. For more information, see the Contact Center Administrator Guide, section Scenario Entries Overview.

  1. Go to Scenario Entries > Dial-in and add a new scenario entry.

  2. At the most basic level, add Name, Scenario, Service, Internal numbers, External numbers, and Transcribe calls (if no STT account is configured, this will be greyed out).

  3. Click Apply.

You're Done

This completes your conversational IVR configuration. Next, you should try it out by calling your contact center's access number and interacting with the IVR and bot. You can always come back to your scenario and entry point for further editing.

Recommended Reading

For more information on inbound voice configuration, bots, integrations, and scenarios, see these Bright Pattern tutorials:


How to Load Voices from Google Cloud Text-to-Speech API

A Google Text to Speech integration account uses your Google Cloud API credentials to connect Bright Pattern Contact Center to Google's TTS engine. Once configured, integration lets you create voice prompts in numerous supported languages to automate bot responses.

Configuring a such an integration account requires that you not only provide credentials for access, but that you also authorize voices in a specific language on the Google end.

This article describes how to do just that and load voices to your Google Text to Speech integration account.

Procedure

Step 1: Go to Integration Account Properties in Contact Center Administrator

  1. In Contact Center Administrator > Call Center Configuration > Integration Accounts, add a new Text to Speech integration account. For property descriptions, see the Contact Center Administrator Guide, section Text To Speech Integration (TTS).

    Select the type of TTS integration account


  2. See the Load Voices property.

    Google TTS integration account properties


  3. To get the list of voices to appear, you need to go into Google Explorer and authorize a list in a specific language to be used.

Step 2: Authorize in Google

  1. Go to https://developers.google.com/apis-explorer/?hl=en_US#search/voices/m/texttospeech/v1/texttospeech.voices.list

  2. In languageCode, enter the language code (e.g., "en-US")

    Enter language code


  3. In fields, enter "voices"

  4. Click Authorize and execute

    Authorize it


  5. In the Select OAuth 2.0 Scopes dialog, select the checkbox for the scope to use

  6. Click Authorize and execute (again)

  7. In the next dialog, allow access to your Google account

  8. This initiates the request (GET list in languageCode) and returns the response code and list (JSON)

    JSON response


Step 3: Load Voices

  1. In Contact Center Administrator, go back to your TTS integration account properties and click Load Voices. This may take a few minutes. When the voices are loaded, a message will display the date they were added: “Loaded on XX/XX/XX.”

  2. Click Apply to save your changes.



How to Load Voices from IBM Cloud Text to Speech API

An IBM Cloud Text to Speech integration account uses your IBM Cloud Text to Speech service credentials to connect Bright Pattern Contact Center to IBM's TTS engine. Once configured, integration lets you create voice prompts in numerous supported languages to automate bot responses.

Configuring a such an integration account requires that you get a list of available voices from IBM using the list voices API method.

This article describes how to do just that and load voices to your Watson Text to Speech integration account.

Procedure

Step 1: Go to Integration Account Properties in Contact Center Administrator

  1. In Contact Center Administrator > Call Center Configuration > Integration Accounts, add a new Text to Speech integration account.

  2. Select Watson as the account type.

    Select the type of TTS integration account


  3. The Text to Speech integration account properties will open.

    Watson IBM Text to Speech integration account properties


  4. Set the URL of your IBM Text to Speech service instance URL. You can find the URL in your Text to Speech service credentials. For example: https://api.us-south.text-to-speech.watson.cloud.ibm.com/instances/0123a4c0-5d67-8cec-9c10-a0cdbb1234b5

  5. Set the API Key, which is found on IBM Cloud in your Text to Speech service credentials.

  6. The Load Voices property is grayed out and you cannot click it until you proceed with Step 2.

Step 2: List Voices in IBM Cloud

This step uses IBM Cloud's List Voices and Get Voices API methods.

  1. In your web browser, go to https://<your-ibm-tts-instance-url>/v1/voices. For example: https://api.us-south.text-to-speech.watson.cloud.ibm.com/instances/0123a4c0-5d67-8cec-9c10-a0cdbb1234b5/v1/voices

  2. Authenticate with your IBM Text to Speech credentials:
    1. Username - apikey

    2. Password - The same Text to Speech API key as set in the integration account properties

      TTS-Watson-Load-Voices-5312.PNG


  3. A list of known models (voices) will be shown in JSON.

  4. You may optionally select the URLs for the individual models and open them in your browser to get more information.

Step 3: Load Voices

  1. Go back to the Contact Center Administrator application, to the Watson TTS integration account properties, and click Load Voices. This may take a few minutes. If the voices field still says "empty" after clicking Load Voices, wait a few more minutes and try again.

  2. When the voices are loaded, a message will display the date they were added: “Loaded on XX/XX/XX.”

    Example of voices loaded on <date>


  3. Click Apply to save your changes.