Chatbots have flooded the corporate travel space in the past
two years. There's artificial intelligence under the hood of each, taking in
traveler preferences, travel policy parameters and historic booking data and
combining it with aggregated content from traditional sources like global
distribution systems and sometimes nontraditional sources like Expedia,
Priceline subsidiaries and Booking.com. Crunching it all together and
delivering personalized itinerary choices in a simple interface is just one
piece of the puzzle for Mezi, and the technology is sharp. Shinde claims Mezi
can process 60 percent of requests without a failure, and should be at 80
percent within six months. When the technology does fail, Mezi has people
behind the scenes ready to take over. That human assist is where Shinde breaks
with tools like HelloGbye or Hello Hipmunk, which strive for 100 percent AI
capability. Instead, it sides with Kayak co-founder Paul English, who has
launched Lola with similar human underpinnings but hasn't fully entered the
corporate market.
Where Shinde has succeeded and competitors trail is in the
partnership business. Mezi hasn't ruled out corporate-direct relationships, but
Shinde's strategy favors travel management companies. Mezi has signed about a
dozen TMC partners, and Shinde scored big-time bringing on corporate travel
tech evangelist Johnny Thorsen this year to pursue those relationships. While
Mezi has additional suitors, Shinde said he's not taking additional partners
right now because his vision is to integrate completely with the ones he has,
not just as a so-called "pocket travel assistant" but as the
next-generation TMC platform. "Our partners need to share that vision,"
he said.
Casto Travel CEO Marc Casto
does. He partnered with Mezi to launch a white-label chatbot, plus an
alternative AI-supported service workflow with specialized agents who are
unfettered by GDS codes and archaic interfaces. He sees a future in which AI
solves an industry labor crisis for qualified agents. "Our AI-assisted
agents are at least five times more productive than a traditional agent,"
he told BTN in September. "And the excitement among [them] has been
amazing."