From Bright Pattern Documentation
Jump to: navigation, search
• 5.3

How to Set up Your Contact Center for Voice Call Transcription

This tutorial will help you to configure your contact center to capture transcripts of call recordings. Turning on transcription is a multi-part process that essentially consists of integrating resources such as Speech to Text (STT), Text to Speech (TTS), and Natural Language Understanding (NLU) from a third-party platform, and enabling the resources to be used in your Bright Pattern Contact Center voice services.

When done, you will be able to view transcripts of recorded calls in Interaction Records, review transcripts for use in Omni QM evaluations, see transcripts in chat from Conversational IVR, and more.

Prerequisites

Transcription requires that your service provider first enable Speech To Text and integration with Text to Speech providers for your contact center. If those features are not enabled, you will not be able to configure the necessary integrations.

Bright Pattern Contact Center supports STT integration with IBM Watson, Google, and GoVivace; TTS integration with IBM Watson and Google; and NLU integration with IBM Watson. You must have accounts with the appropriate providers.

Procedure

Step 1: Set up STT

If using IBM’s Speech to Text service:

  1. Sign in to your IBM Cloud account. If you do not already have an IBM Cloud account, sign up at https://www.ibm.com/cloud.

  2. Go to Dashboard > Resource List (https://cloud.ibm.com/resources) and click Create resource.

  3. Search the catalog and add the Speech to Text resource.

  4. Set up Speech to Text by selecting the region closest to you and choosing a plan.

  5. On the Manage page, copy the API key and the URL.

  6. In Bright Pattern’s Contact Center Administrator application, section Call Center Configuration > Integration Accounts, add a Speech to Text integration account of type Watson.

  7. In the integration account properties, do the following:

    1. Paste your URL (the one you copied from the service), and change the "https" to "wss" so the format is as shown: wss://api.us-south.speech-to-text.watson.cloud.ibm.com/instances/1a23456f-121d-4c52-bc06-62168f5a18de

    2. Paste your API key (the one you copied).

    3. Select API key.

    4. Select default account.

    5. Click Test Connection to make sure the connection to IBM Speech to Text is valid.

    6. Click Apply to save your changes.

If using Google Cloud Speech-to-Text:

  1. Sign in to the Google Cloud Platform. If you do not already have a Google Cloud account, sign up at https://cloud.google.com/.

  2. From your Home console, go to APIs and Services > Dashboard and click Enable APIs and Services.

  3. Search the catalog and add Cloud Speech-to-Text API.

  4. Click Enable API.

  5. In Credentials, copy the API key and the URL.

  6. In Bright Pattern’s Contact Center Administrator application, section Call Center Configuration > Integration Accounts, add a Speech to Text integration account of type Google synchronous STT.

  7. In the integration account properties, do the following:

    1. Name the account.

    2. Paste your API key (the one you copied from Google Cloud).

    3. Click Test Connection to make sure the connection to Google Cloud is valid.

    4. Click Apply to save your changes.

If using GoVivace:

  1. Sign in to your GoVivace account. If you do not already have an GoVivace account, sign up at https://www.govivace.com.

  2. Go to GoVivace’s Speech to Text software and copy the API key.

  3. In Bright Pattern’s Contact Center Administrator application, section Call Center Configuration > Integration Accounts, add a Speech to Text integration account of type GoVivace.

  4. In the integration account properties, do the following:

    1. Name the account.

    2. Set the URL as wss://services.govivace.com/telephony

    3. Paste your API key from GoVivace’s Speech to Text software.

    4. Select default account.

    5. Click Apply to save your changes.

Step 2: Set up TTS

If using IBM’s Text to Speech service:

  1. Sign in to your IBM Cloud account. If you do not already have an IBM Cloud account, sign up at https://www.ibm.com/cloud.

  2. Go to Dashboard > Resource List (https://cloud.ibm.com/resources) and click Create resource.

  3. Search the catalog and add the Text to Speech resource.

  4. Set up Text to Speech by selecting the region closest to you and choosing a plan.

  5. On the Manage page, copy the API key and the URL.

  6. In Bright Pattern’s Contact Center Administrator application, section Call Center Configuration > Integration Accounts, add a Text to Speech integration account of type Watson.

  7. In the integration account properties, do the following:

    1. Paste your URL (the one you copied from the service): https://api.us-south.text-to-speech.watson.cloud.ibm.com/instances/ab12a3f4-a567-8910-b123-4b5a67891abc

    2. Paste your API key (the one you copied).

    3. Click Load Voices.

    4. Click Test Connection to make sure the connection to IBM Text to Speech is valid.

    5. Click Apply to save your changes.

If using Google Cloud Text-to-Speech:

  1. Sign in to the Google Cloud Platform. If you do not already have a Google Cloud account, sign up at https://cloud.google.com/.

  2. From your Home console, go to APIs and Services > Dashboard and click Enable APIs and Services.

  3. Search the catalog and add Cloud Text-to-Speech API.

  4. Click Enable API.

  5. In Credentials, copy the API key and the URL.

  6. In Bright Pattern’s Contact Center Administrator application, section Call Center Configuration > Integration Accounts, add a Text To Speech integration account of type Google TTS.

  7. In the integration account properties, do the following:

    1. Name the account.

    2. Paste your API key that you created for your Google Cloud APIs and services. Note that this is not the same as the service key ID.

    3. Click Load Voices. Follow these instructions to load voices.

    4. Click Test Connection to make sure the connection to Google Cloud is valid.

    5. Click Apply to save your changes.

Step 3: Set up NLU

  1. Sign in to your IBM Cloud account. Go to Dashboard > Resource List (https://cloud.ibm.com/resources) and click Create resource.

  2. Search the catalog and add the Natural Language Understanding resource.

  3. Set up Natural Language Understanding by selecting the region closest to you and choosing a plan.

  4. On the Manage page, copy the API key and the URL.

  5. In Bright Pattern’s Contact Center Administrator application, section Call Center Configuration > Integration Accounts, add a Watson Natural Language Understanding integration account.

  6. In the integration account properties, do the following:

    1. Name the account.

    2. Paste your URL.

    3. Set your authentication method as API key, and then paste your API key.

    4. If you are using a legacy Watson NLU account, set authentication method username/password (legacy) and type your username and API key. If you don’t know your username or password, you can set username “apikey” and set the password as your actual API key.

    5. Set your limit (e.g., 100, 1000, etc.).

    6. Click Test Connection to make sure the connection to IBM Speech to Text is valid.

    7. Click Apply to save your changes.

Step 4: Set up your service

  1. In Contact Center Administrator, section Call Center Configuration > Services & Campaigns, do the following:

    1. Select the voice service that you want to have transcription.

    2. Select the checkbox for the following:

      1. Record calls

      2. Record IVR call segment

      3. Transcribe calls

    3. Select your new NLU account under Natural Language Understanding.

    4. Click Apply to save your changes.

  2. In section Scenario Entries, make sure you select the service that you just set up for call recording and transcription. Click Apply to save your changes.

Step 5: Test by making a call

Log in to the Agent Desktop application and accept a service call. After the call is completed, wait about 15 minutes for the recording to be aggregated, and then check Interaction Records for the transcript.

< Previous