Check out the on-demand sessions from the Low-Code/No-Code Summit to learn how to successfully innovate and achieve efficiency by upskilling and scaling citizen developers. Watch now.


Over the past few years, we’ve all encountered “Let’s chat!” buttons on websites that promise a quick, helpful customer service experience. But heavily hyped AI-driven chatbots, an important part of the customer experience mix since 2016, have also proven to be a mixed bag. Consumers found many bot interactions disappointing and time-consuming. Meanwhile, enterprises often needed to provide far more costly care and feeding of chatbots than expected. 

Thanks to open-source AI language models such as Google’s BERT and Open AI’s GPT, it’s now far easier for organizations and technology software vendors to build on top of these innovations. They can create more sophisticated conversational AI tools, from smarter chatbots and asynchronous messaging to voice and mobile assistants. These days, deep learning models can be designed quickly. And, depending on how they’re done, they might need only a small amount of training data, Hayley Sutherland, senior research analyst for conversational AI at IDC, told VentureBeat. 

“Over the last two or three years, the ability for machines to understand both written and spoken human language has really, really improved,” she said. “The technology has moved quite a bit past what we would think of as a rules-based, scripted approach where a human manually writes a rigid script that, if it goes outside of that, can break easily and increase frustration.” 

Now, machines can not only better understand the words being said, but the intent behind them, while also being more flexible with responses. “That means we can create much more sophisticated virtual assistants or customer care agents, whether they are text-based or voice-based,” Sutherland said. 

Event

Intelligent Security Summit

Learn the critical role of AI & ML in cybersecurity and industry specific case studies on December 8. Register for your free pass today.

Register Now

Solving for conversational customer challenges

Derek Roberti is VP of technology, North America at Cognigy, a low-code conversational AI platform based in Germany. It was named a leader in Gartner’s Magic Quadrant for enterprise conversational AI platforms, along with companies such as Kore.ai, OneReach.ai and Amelia. According to Gartner, the enterprise conversational AI platform market “focuses on the needs of larger enterprises by targeting multiple use cases, modalities of conversation (such as speech, chat, text messaging and email) and the ability to operationalize within the enterprise.” 

Overall, the conversational AI market in the customer service space is divided into three key categories, Roberti explained. The first are conversational AI specialists, with platforms that have user interfaces tailored for both the technical and non-technical user; out-of-the-box integrations; and a wide variety of channels. “Those are the ones that Gartner has called out as leaders in the space,” he said. 

Next, there are the giant cloud providers, such as Microsoft, Google, Amazon and IBM. “These offer core services, perhaps translation, or natural language understanding, or speech-to-text, but don’t necessarily have that set of user interfaces and prebuilt components,” he added. Finally, there are also thousands of other, smaller market players taking advantage of open source innovations to provide off-the-shelf tools with varied levels of sophistication.  

All of these companies, across categories, are “working to solve the same problem,” said Roberti. That is, to create first-class customer experiences, particularly with tooling accessible to both the non-technical and the technical builder. “How can we empower people to build automated interactions that are welcoming, easy to get started with and lets you build out even the most advanced conversations?” he explained. 

Conversational AI targets two types of customer service buyers

Roberti cites two primary types of buyers in the market for conversational AI tools for customer service and support. First, there are buyers who own the contact center or customer-facing support systems. “These are generally non-technical buyers,” he said. 

These buyers may have never worked with conversational AI before, or don’t have developer resources, Sutherland added. “That’s why we’re increasingly seeing these kinds of low-code or no-code tools,” she said. “You can have someone who isn’t a developer but is an expert in customer conversations – who knows what a good conversation looks like, who can help to train and check on the capabilities of that conversational AI that’s being built and really ensure that the human element is there.” 

On the other hand, there are more technical buyers, including enterprise architects, who get requests from every part of the organization for chatbot and voice automation capabilities, Roberti explained. “They’re looking for a platform that can be used across the enterprise. They do care about the user interfaces, but they also care about how the tools will integrate into other systems and how it works within security and compliance ecosystems.” 

Context-aware conversational AI is essential 

Quiq is a Bozeman, Montana-based AI-powered conversational platform that enables brands to engage customers on the most popular asynchronous text messaging channels. According to founder and CEO Mike Myer, first-generation chatbots lacked good natural language capabilities and often did not allow customers to access the right data. 

They also had very little context awareness to boost personalization, Myer explained: “For example, If you recently completed a purchase and a couple of days later you come back to the website, it’s helpful if the chat box actually says ‘Welcome back. Do you have a question about the order you placed yesterday?’” 

Now, however, conversational AI technologies, such as the underlying AI-driven natural language capabilities, are reaching a plateau. “The difference between vendors when it comes to natural language understanding is imperceptible from a customer perspective,” he explained.  “What is different now is the quality of implementation, the design, how much training has gone into it.” The UX, he said, “has become the big differentiator.” 

Acquisitions lead to holistic conversational offerings

It’s a sign of the massive, fragmented conversational AI market in the customer service space, as well as the VC money flowing into it, that Sutherland told VentureBeat that she had not heard of Quiq. That is even though the company recently announced a $25 million series C funding round and last year acquired Snaps, another conversational AI tool. 

“That’s very characteristic of this space right now, the really big infusions of VC money,” Sutherland said. “We really are at an inflection point where we’ll start to hit some consolidation.” 

In fact, acquisitions have become a regular occurrence in the space. Beerud Sheth, cofounder and CEO of messaging leader Gupshup, recently announced three conversational AI acquisitions, including Active.ai and AskSkid, while adding, there are another two in the pipeline. 

“We constantly evaluate in terms of technology, so if we find some other company that’s done something interesting that augments what we do we will happily consider it,” Sheth said, explaining that the acquisitions become part of a holistic platform, as conversational AI becomes part of what every business will need. 

Sutherland also says a smaller conversational AI company, Uniphore, is making interesting acquisitions to round out their AI-driven offerings. “One was a company using AI to analyze video and help salespeople understand customer sentiment,” she explained. “At the time, Uniphore was mainly focused on customer care, but now there’s this sales-focused conversational AI. The idea is that companies might eye the rest of the customer funnel.” 

Tight labor market leads to smarter conversational AI 

A tight labor market is driving conversational AI growth in customer service, said Roberti. In the early days of the pandemic, many contact center agents were let go, for example. “Now, even if they could hire as many people as they needed to at 25% more in compensation, staffing is not available,” he said. “So companies are being pushed towards automation as an imperative, as a matter of survival.”  

The good news is, the latest in conversational AI for customer service has the potential to improve the image of an industry previously filled with unhelpful chatbots, Roberti said: “I would say if you did a customer satisfaction survey about chatbots and voice bots in January of this year and compare it to January 2023, you will see a much more favorable reaction.” 

VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.