IBM Cloud Satellite: Run and manage services anywhere Learn more

Create a next-generation call center with Voice Agent

This article is part of the Watson Assistant learning path. See the Watson Assistant page for more information on features and getting started.

Summary

IBM Voice Agent with Watson provides a way to integrate a set of orchestrated Watson services with a public or private telephone network by using the Session Initiation Protocol (SIP). This code pattern walks you through a Voice Agent example that enables you to automate your call centers using Watson Assistant, Watson Speech to Text, and Watson Text to Speech, without having to orchestrate between all of the different services yourself. The Voice Agent does that for you.

Description

IBM Voice Agent with Watson enables direct voice interactions over a telephone with an artificial intelligence (AI) self-service agent. IBM Voice Agent with Watson lets you:

  • Improve telephone-based customer service: Connect IBM Watson services to your telephone systems, allowing it to answer queries from callers, perform call triage, and transfer the call to the appropriate agents.
  • Drive down costs: Reduce costs by using IBM Watson to handle some of your calls instead of using call center agents.
  • Deployment flexibility: Connect seamlessly to IBM Watson services in IBM Cloud from either the cloud-based Voice Agent with Watson or on-premises with IBM Voice Gateway.

When you have completed this code pattern, you will understand how to:

  • Provision Watson Assistant, Speech to Text, and Text to Speech services on IBM Cloud
  • Import a sample call center dialog to Watson Assistant
  • Connect Voice Agent to Twilio using the SIP communication protocol
  • Improve speech recognition accuracy by creating a Watson Speech to Text custom language model with a grammar (optional)

Flow

flow

  1. User phones a call center phone number that is associated with a Twilio account.
  2. The Twilio number connects to a configured Voice Agent service on IBM Cloud by using the SIP communication protocol.
  3. Voice Agent calls Watson Speech to Text to transcribe user input.
  4. Based on user input, certain flows in Watson Assistant are triggered. As required, the flow switches between a general purpose or customized Speech to Text model.
  5. Text responses from Watson Assistant are transmitted back to the Voice Agent service by using Watson Text to Speech.
  6. Voice Agent sends the audio back by using the Twilio Elastic SIP Trunk.
  7. The resulting audio is played back to the user.

Instructions

Find the detailed instructions in the README. These steps explain how to:

  1. Deploy the Voice Agent on IBM Cloud.
  2. Create Watson services on IBM Cloud.
  3. Configure Voice Agent and Twilio.
  4. Import the dialog skill into Watson Assistant.
  5. Create a Speech to Text custom language model.

Conclusion

This code pattern walked you through a Voice Agent example that enables you to automate your call centers using Watson Assistant, Watson Speech to Text, and Watson Text to Speech, without having to orchestrate between all of the different services yourself. The code pattern is the final part of the Watson Assistant learning path. You should now have a better understanding of Watson Assistant and its features. If you want to learn even more about Watson Assistant, take a look at the Watson Assistant product page.