Voice Tech Leans on AI to Humanize Customer Service and Experience

If you find chat bots unhelpful or holding for customer service reps a drag, don’t feel bad. It’s a natural human reaction — the kind of human reaction better handled by automated voice assistants powered by artificial intelligence (AI) that are smart, fast and far more humanlike now.

Companies are automating everything imaginable for cost savings, efficiency and accuracy in the pandemic era. However, the installed base of not-too-bright chat bots and the inconsistent voice systems they commonly use are quickly becoming outmoded, and customer sentiment is taking a bruising in the process.

This state of affairs is fast-tracking consumer experience (CX) to the next level of automated interaction: deploying voice systems that learn and improve, rapidly tapping vast datasets, and creating chat bots and automated voice helpers that are not just better — they’re even empathetic.

It’s the arrival of true conversational commerce. Raghu Ravinutala, CEO of customer experience automation firm yellow.ai, described the startling advances in conversational commerce in a recent human-to-human conversation with Karen Webster that centered in dynamic AI agents.

Listing phases that automated customer service passed through in recent years, Ravinutala said, “The next [stage] of evolution is more cognitive understanding and doing cognitive tasks like negotiating [and] assuaging the customer.

“That’s the next level of automation that we are seeing that gets closer to humans. I think the AI models are there, but the data about what kind of empathetic responses are delivered, what kind of outcomes, this data is still being collected.”

Data collection and integration is “the biggest bottleneck” in dynamic AI’s advance, he said, and vital “to train these models to elicit the right behavior from these agents for end consumers.”

Ravinutala is referring to the dynamic AI solutions his company specializes in.

“These are AI agents that are intricately integrated with core enterprise systems,” he said. “For example, somebody is looking to reschedule an airline ticket. The ability to integrate with the billing system and inventory [customer relationship management system] and get workflows done is a core component of dynamic AI agents.”

Observing that communication with customers is not unilaterally inbound or outbound, nor unilaterally voice or chat, “the ability of these AI agents to choose the right channel and the right mode of communication with the end customers is a key element,” Ravinutala told Webster. “That’s why we call them dynamic.”

Ravinutala added that dynamic AI agents are inclusive, saying, “We understand that consumers of global companies talk in different languages, and they move across different channels of communication.

“The ability of agents to naturally adapt to the different language of communication and the right channel at the right point of time, these are all needed to address global consumers and the changing complexities of engaging with [them].”

See also: Voice Operating Systems Like Alexa Will Power the Connected Economy: Here’s Why

Data Drives Automated Experiences

As to whether brands and retailers using last-gen automation are aware that expectations around CX have shifted drastically in the past 20 months or so, Ravinutala believes that realization is happening, and some have already pulled ahead with it.

Citing Sephora’s reputation for personalized in-store experiences, he said, “The brand recognized that they need to provide similar experiences while [people] are interacting on a digital medium. So, they have taken all the data products and they have enabled an AI agent for end customers who can get product recommendations based on their needs.”

A change in corporate mindset is part of the challenge facing wide adoption of dynamic AI agents, as Ravinutala feels companies have been “forcing” bad automation onto good customers.

Saying “there are areas where machines and automation are superior to human-based interactions,” Ravinutala noted that “automation and machines are not evolved to a level where they can provide empathy.”

Using the example of the air passenger who needs some comforting COVID-19 support, he added, “You really need a human to provide that assurance and provide that empathy.”

“Clearly, a model where a dynamic AI agent can switch the context between an automation and a human providing the end customer experience seamlessly without them needing to repeat the information is the solution that the pioneering enterprises are adopting,” he continued.

See also: Conversational Commerce May Be ‘Significant Portion’ of eCommerce Purchases Within Next Decade

Chatbots, Voice and Video Merging

Advances in both chatbots and voice recognition technology are hardly going unnoticed. As digital commerce grows and the pandemic wears on, the concept is getting real traction.

After stores were suddenly closed in 2020, many retailers and brands found themselves cut off from customers. Many caught on to other messaging channels, and it goes out from there.

“The immediate medium to communicate with the end customers was messaging because that’s the property that’s on everybody’s phones and natural to the users,” Ravinutala said. “Now brands like Adidas, brands like Sephora, Domino’s, [and] Procter & Gamble have started experimenting and providing engagement with the end consumers directly on these messaging platforms.”

He added that “platforms are delivering higher engagement than any other digital property they ever had, from mobile apps to websites because it’s so natural.”

“They’re on the contract list of the end consumer,” Ravinutala continued. “The second thing that they kind of learned is that the messaging is not just supplementing the existing digital properties, but you can drive end-to-end transactions on these messaging platforms.”

Adidas, for instance, “now actually sells shoes on messaging apps,” he said. But for give-and-take data, nothing beats a conversation — even an automated one.

“Conversations are absolutely specific,” Ravinutala said. “‘I need this, I don’t need this.’ They are getting much, much deeper insights [into] end customers through conversations rather than websites or mobile app-based interfaces. This is leading them to build messaging and conversations as a separate channel for customer acquisition and customer engagement.”

From here, Ravinutala sees digital assistants like Alexa and Google Assistant replicating a type of store.

“I believe that voice will move from analog voice that’s on telephony to more digital voice, because a lot of platforms like Apple FaceTime, Apple messaging, all of these have formats where you can actually make audio calls with the end consumer,” Ravinutala said. “So, voice will move to digital and will evolve into video. I see the future of brand interactions with consumers as virtual video interactions as the starting point.”

See also: Improved AI Gives Chatbots a Second Life in the Digital Ecosystem