AI Adoption in Contact Centres

Mike Palmer at Spearline explores how the voicebot is emerging as an attractive conversational user interface which offers accessible and comfortable user interaction.

These can assist users in a range of activities, including access to product and service information over telephone or over the internet.

From humble roots in simple rules-based interactions and robotic voice, the underlying tech now leverages artificial intelligence (AI) and advanced voice analytics to support a more dynamic and more complex exchange with the user.

Impressive implementations offer realistic mimicking of human-to-human conversation. Contact centre teams are enthusiastically rolling out virtual agents.

According to V-Soft Consulting, before the emergence of today’s voicebots, voice technology existed in the form of voice recognition systems, but these systems failed terribly. This is because the mechanisms either did not integrate with corporate systems or did not provide enterprise-grade reliability.

AI has driven huge improvements and is the fundamental differentiation when compared to earlier speech systems.

The global market for voice technologies is estimated to grow to $184.9 billion in 2021. In the U.S. more than one-fourth of organizations are using intelligent agents and the rest are planning to invest in this technology in the future.

The rise of voice assistants in the home has driven a level of comfort for many consumers that opens the door wider for virtual agents.

The idea of voice assistants, surprisingly, is not new. An idea that has been around since the ’60s, humans are constantly trying to make artificial intelligence and conversational systems interaction a reality.

As a business application, voicebots can:

• Improve workplace service relationships. Organizations have started using chatbots and voicebots with service delivery platforms like ServiceNow.
• Pull up info immediately during boardroom meetings. Voicebots can pull info from databases with a simple voice command.
• Monitor IT and operational systems in real time. Notifications can be delivered verbally in natural language.
• Raise service tickets and report status with a voice command. Voice acts as a platform and service enabler.
• Remove the need for hefty dashboards by pulling up appropriate information via voice command.
• Run your entire office administration (HR, workforce management, project management, and so on) with minimal issues.
• Work with any legacy or cloud-based system makes it more adaptable.
• Externally facing customer service – Voicebots have proven their worth by serving any number of customers at any given time, thereby saving human hours and costs.
• Improved usability which is more intuitive, and which can be multilingual.
• Provides for touch-less and screen-less UX where desired

In the contact center arena, customer interactions will largely be via a website, or via a telephone call. In these circumstances, developers face the challenge of voice quality in the communications channel impacting the performance of the application.

In a recently published whitepaper, “Leveraging voice technology”, Spearline examined the relationship between audio channel quality and word error rate (WER) on leading voice technologies. As quality degrades the success of the voice technology falls significantly.

Sophisticated developers will build more “conversational” error detection and correction into their applications, further mimicking our human capacities. In response to an uncertain input, the voicebot will branch to “Did you say x?” or “I’m sorry, can you say that again?”

Of course the effectiveness of programmed error handling will be hindered by persistent issues with the audio channel. Proactive monitoring of voice services, SIP trunks and network arrangements can ensure a solid environment for successful virtual agent deployment.