AI travel

A Mobi platform is available that leverages multiple artificial intelligence (AI) models to surface recommendations that promise to improve any travel experience.

The platform is based on a collaborative approach that makes use of AI models, including large language models (LLMs) and visual language models, to combine natural language text, image, numerical and categorical data in way that surfaces recommendations based on the known interests of the traveler. Developed at the Massachusetts Institute of Technology (MIT) for more than a decade, the platform can, for example, surface additional points of interest for a traveler based on their known interests.

A business traveler with an interest in history, for example, would be informed of interesting sites to visit that are nearby where they already plan to travel.

The goal is to leverage the various AI technologies for the right task to provide a better travel experience, says Mobi CEO Anna Jaffe. “We want to be hyper human centric,” she says.

At the same time, the platform can also be used by providers of travel services to maximize profit and revenue based on itineraries.

Travel companies that have partnered with Mobi include Marriott, TUI Group and the Singapore Tourism Board.

There are currently three iterations of the Mobi platform that provide user interfaces for travelers, concierges and travel agents. In effect, the Mobi platform is using AI models optimized for routing to enable travelers make optimal use of their time, notes Jaffe.

Providers of travel services have been using various technologies to try and route passengers optimally for decades. The platform created by MIT is taking that to the next level by making it easier to personalize itineraries involving multiple providers of travel services, including airlines, trains, hotels, rental car agencies, tourist destinations and restaurants.

There’s no doubt many providers of these services have already launched multiple AI models but the Mobi platform aggregates signals from multiple sources to provide a potentially more optimal experience. The more providers of travel services that participate the more data there is to generate an optimal set of recommendations for each traveler.

There, of course, might be privacy concerns, but most of the data that Mobi requires has already been shared with a travel services provider. Individuals can either opt into the service directly or give their travel services provider permission to use their data. Many travelers are already using generative AI platforms such as ChatGPT to plan itineraries. The Mobi platform is looking to take planning to the next level by applying routing analytics to provide a more optimal experience.

It’s still early days so far as applying AI to travel is concerned, but given everything from reducing carbon emissions to reducing travel costs there is plenty of opportunity to optimize travel, especially during busy travel seasons such as Christmas. The challenge and the opportunity now is to convince the providers of all those services to collaborate in a way that benefits travelers as much as the provider of the travel service.