AI and machine learning 'poised to accelerate' pace of change
Sabre tips new technologies to transform travel retailing
Artificial intelligence and machine learning are poised to transform travel technology and to do so at an accelerating pace.
That is according to Sergey Shebalov, vice-president and head of research at Sabre Labs, part of US‑based travel technology provider and global distribution system Sabre.
Shebalov leads a research team of about 70 working on the introduction of machine learning on a platform hosted by Google.
He told Travel Weekly: “The technology is changing drastically. Our data and systems are migrating to Google, which provides artificial intelligence [AI] and machine learning [ML] capabilities that allow our data scientists to build ML applications. We see a real difference to how we worked before.”
Sabre signed a 10-year partnership with Google Cloud in January 2020 and the platform was launched last year.
One immediate change, he said, is “the ability to manage huge amounts of data. It is much more convenient to grab data, to manage data, to train a machine-learning platform and to see patterns. AI and ML are very good at digging through a vast amount of data and this allows development of a lot of tools.
“For example, to be able to construct a reliable [fare] offer to a customer, we need to know their preferences, their attributes, and we need to recognise different segments.
“For retailing, we need to determine what the price should be for an offer, not just the flight but the ancillaries, and this needs to be very fast.”
The technology “takes all this into account, including the strategy of the airline on pricing”, he said.
But that is not all. Shebalov explained: “We must display an offer the customer is likely to pick up. That is particularly important with mobile because only two or three offers display [on a screen]. It’s important the offers are relevant.” In addition, “for a travel agency or OTA, the display must look convenient”.
That is just the retailing. AI and ML can also transform ‘fulfilment’, he said, for example during a shutdown of traffic.
“When informing customers or reimbursing them, we need to look at all available seats and we want the options in minutes. It’s impossible manually. You need very complicated algorithms. AI allows us to solve these problems.”