Talk to the bot: How conversational AI is changing the world

Johannesburg, 28 Jan 2021
Read time 3min 30sec
Sebastian Reeve, Director of International Go-To-Market, Intelligent Engagement, Nuance Communications
Sebastian Reeve, Director of International Go-To-Market, Intelligent Engagement, Nuance Communications

It’s next to impossible to interact in the modern, digital world without interacting with artificial intelligence (AI). As we lean on technology to simplify our lives, so our use of AI, often in ways that are almost invisible, continues to increase.

If you think about many of our day-to-day activities, from checking your favourite weather app, to what ads you see online; from speaking to your choice of digital assistant to getting information from your bank via WhatsApp, all of these are powered by AI.

The ubiquity of AI underplays the complexity of the engineering that goes into these systems, explains Sebastian Reeve, Director of International Go-To-Market, Intelligent Engagement at Nuance Communications.

“The AI that powers modern technology isn’t something that simply emerged one day, it’s the culmination of decades of work that began with the creation of speech recognition and natural language technologies. These technologies, many of them developed by Nuance, power the interactive voice response systems that you use each time you call into a contact centre and provide the foundation of today’s smart digital assistants.”

Conversational AI and omnichannel digital engagement

Whether you’re chatting to a virtual assistant online or over the phone, conversational AI is what allows it to hold a human-like conversation. “This is at the heart of our own virtual assistant, Nina, which has been deployed by everyone from H&M and Swedbank, to the Australian Tax Office, allowing them to ‘speak’ to people on their own terms,” he says.

He adds that conversational AI is not just the most prominent way that ordinary people encounter AI, but it’s also one of the most complex. “Even today we’re still at a point where even the best virtual assistant can’t meet the digital needs of today’s customers on its own.

“As digital channels have evolved, most brands have added point solutions over time,” says Reeve. “Too often, the result has been costly, siloed operations and fragmented customer experiences. To resolve this issue, companies are investing in unified digital engagement platforms that bring together virtual assistants, live chat and multi-channel customer messaging and proactive notifications, as well as agent tools and contact centre analytics.”

The use of AI voice recognition is a critical component of the next generation of authentication services for customer service applications. Customer service agents have, for years, struggled with myriad increasingly complex authentication methods. While these have proved effective in the past, the growing dark Web market for personally identifiable information means that even this is no longer a foolproof method of authentication.

“Leveraging AI, we’ve created biometric security solutions that include technologies that are able to identify customers by the sound of their voice, the language they use, and even how they hold and interact with a mobile device,” explains Reeve. “We’ve seen how this has enabled companies to cut 15% off their average call times and even identify, prioritise and protect elderly customers during the pandemic.”

“Going beyond simple authentication it’s possible to leverage AI-based voice recognition systems to bolster fraud prevention. Just like banks in the old days used to have pictures of known criminals on the wall, today they can identify criminals by matching their voice to a database of known career criminals,” he says.

While the use of conversational AI has progressed massively from its early days, Reeve comments that there is still room for improvement. “With people looking to make the most of their free time, they’re going to look towards conversational AI as the method of choice for many everyday tasks, requiring greater sophistication and security from the underlying technology.”