Corporate booking tool provider Serko has partnered with artificial
intelligence-driven customer service company FaceMe to deliver a human-like travel
digital assistant that responds verbally to travel requests. The technology
behind the "chat" side of the chatbot is similar to what is already
found in Zeno's Ask Zeno platform, with the same content, the same policy
configurations and the same AI-based workflows and user recognition. But it
delivers information via an interface that simulates human facial structure and
expressions associated with listening and reacting to requests.
That's where FaceMe literally enters the picture. The New
Zealand-based AI customer service platform has created the visual "bot"
part of the chatbot. The company offers hundreds of different "looks"
in terms of visualizing male, female or neutral bots and it can build bots to
spec—to look like a lead consultant at the agency or other customizations that
communicate a company's brand or culture. So, it looks like an agent; does it offer
service like one? Judge for yourself.
Serko Zeno Digital Assistant
Serko CEO Darrin Grafton told BTN the company is
looking to establish with corporate travelers an "emotional relationship
to the technology" or "to the brand," that makes people want to
come back and use it again. He posited it's a connection that goes beyond
what a flat chat interface can offer. "We see this as an extension of the
tools already in play: online booking tools, mobile
booking tools, chat and now we have a digital assistant."
To that end, the technology offers an enhancement of the
chat interface. Not only has Serko attempted to simulate the appearance of
human emotions with bot visuals, it also has used tools like Microsoft Cognitive
Services to enable the digital assistant to recognize user voice and facial
cues via microphone and webcam to detect when the traveler is frustrated with the
system or confused. At that point, said Grafton, the system can flip to a human
agent to resolve issues.
For now, he said, the digital assistant works on simpler
interactions like making and changing bookings, but the
conversational interface holds promise for trickier decision-making workflows that
may be required sooner rather than later.
Take New Distribution Capability. The Zeno booking tool is
built to accommodate NDC's complexity, as demonstrated at BTN's Innovate
conference last year. However, criticisms levied against the tool included the
number of menu items and drop-downs required to get to final booking through
the NDC workflow. AI-powered conversational interfaces could take a different
approach, Grafton said, leveraging the user's historic data to ask or remind
them of options that might be more efficient or offer more value.
Grafton invoked baggage fees as a prime example of where an
AI-powered bot might suggest better alternatives, by not only using historical travel data but also pulling in expense data to see if bag carriage or any
ancillary service wasn't purchased upfront but added after the booking. With
this "knowledge," the
conversational interface could suggest that the traveler accept a value-based
bundle that includes their past ancillary purchases, eliminating workflow steps
that today must be included in display-based booking tools. The ability to leverage
expense data isn't currently available but is something Serko is looking to roll
out in the next six to nine months.
Could this be accomplished in a flat chat-based platform?
Theoretically, yes. But the human engagement and trust factor might be reduced
compared to the digital assistant. The eagerness of travelers to adopt this type of interface, based on a test group at Flight Centre Travel Group's Illuminate conference last September, surprised Grafton. "You're always trying to balance how fast
to push this type of technology and to soften the process," he said, "but overwhelmingly, the audience said they would use it."
That said, Zeno's digital assistant "doesn't have to
fit everybody," Grafton acknowledged. It's a separate module that comes at
a cost to travel management companies that use Zeno technology. "A TMC may choose to open the opportunity
to a certain set of clients or customize something for a particular
client."
Data security has been a particular concern as Serko developed
the digital assistant technology, given the ability to capture facial
expressions and voice nuances and associate those with the individual user.
Grafton underscored that such data was destroyed after each interaction with
the bot, retaining only what is necessary to feed the user's dynamic archive of
historic bookings, which is what drives the AI.
Top of mind for Grafton is pushing through the limitations
of traditional corporate booking technologies and delivering faster, better and
more engaging results for the user. "We don't have to have the boring
style. Enterprise can have consumer cool with all the business principles that
are needed."