TUI Takes its Data to Exciting New Destinations With Snowflake and Cortex AI

Travel and tourism are a marvel of engineering. Resorts host countless guests, airlines ship hundreds of people to exciting places, and companies like TUI Group use extensive data to match travellers with the ideal destinations. 

With five airlines, 400 hotels, 1600 travel agencies and even 16 cruise ships, TUI Group is the largest leisure, travel and tourism company in the world. Like any large enterprise, TUI wanted a way to unify data across its business.

TUI’s Technology Analytics team is responsible for the pipelines, infrastructure and solutions that support all data in the enterprise. To solve its data challenges, the team deployed Snowflake’s AI Data Cloud, running on Amazon Web Services (AWS). Now, it has a single unified platform for all TUI data — and it has also laid a foundation for AI-powered data search and discovery that will help streamline operations, empower teams to meet higher customer expectations and increase delivery capacity of data teams.

TUI’s airline isn’t the only part of the business that’s in the clouds. The travel company has always made the most of as-a-service systems and infrastructure to empower data users. However, as every team in TUI has unique needs, each business unit was responsible for managing its own data use cases. 

This presented several challenges. It often meant that new hires and those in non-technical roles were overwhelmed by extensive analytics and data tools, with little guidance on how best to use them. And even for those that were familiar with TUI’s technology stack, it wasn’t always obvious how to make the most of its systems to more effectively analyze data, derive insights and engage with the data products.

Anastasiia Stefanska, Senior Data Analyst at TUI and Snowflake Data Superhero 2025 Cohort, had a different approach in mind: “We wanted to be at the forefront of innovation, using a platform that could offer new data features and put us on the cutting edge of AI capabilities. So, I suggested we look at Snowflake.” 

Alina Vishniakova, Head of Technology, Technology Analytics at TUI, says she didn’t need much convincing. “I saw that with Snowflake, we could better support our teams, keep up with new data innovations and get better quality insights across TUI.”

Initially, TUI deployed Snowflake’s AI Data Cloud to accelerate and scale its analytics and improve customer understanding across regions. Since then, Snowflake has become the analytics backbone for everything TUI does with data, whether that’s optimizing flight and hotel occupancy, crafting personalized travel offers or informing internal decisions and business strategy.

Over time though, TUI’s Technology Analytics team has come to see that Snowflake can offer even more. For example, Snowflake’s heavy investment in its Cortex AI tools was a draw. “That opened our eyes to how we could use Snowflake to create AI use cases faster than if we looked at other providers,” says Vishniakova.

Cortex AI now powers TUI’s Technology Analytics Companion (TACO), an AI platform powered by Cortex Analyst that allows users to query data through natural language. With TACO, TUI democratizes analytics, empowers non-technical users and accelerates complex data queries. 

TACO is a powerful system that improves data discovery while offering more insight to more TUI team members than ever before. But it was also an early sign for Stefanska that Snowflake could power TUI’s AI ambitions: “I took the data schema that we had, and asked Cortex to create a semantic model around it and within an hour of tweaking I was getting answers. It was so quick and easy to start working with Cortex.”

TUI also built another AI tool to improve data discoverability. Using Cortex Search, TUI has been able to index all its metadata, making it searchable and discoverable for everyone in the business.

Together, these two use cases have transformed the way users can search for new data products and insights. In the earliest stages of user acceptance testing, over 500 complex data queries were answered in natural language, offering an early sign of the scalability of TUI’s AI use cases. 

Now, every business unit within TUI is experimenting with Cortex AI to make the most of AI tools and large language models. Once these experiments lead to concrete AI tools, the potential is vast, with the Technology Analytics team estimating these use cases could save over 30 days—or even months—of people hours by streamlining processes and empowering teams.

The major benefit of these early AI tools is the speed at which TUI teams can find the data they need and query it for insights. And thanks to TACO’s support for natural language, users can get rapid answers whether they have technical expertise or not. “I can barely keep up with all the new conversations teams are having around data,” says Stefanska. “It’s really exciting seeing things that were previously obscured become intuitive for so many of our users.”

However, TUI has also seen additional benefits from its AI journey with Snowflake, on AWS. While it was indexing data with Cortex Search, the team surfaced and solved data quality issues that had been limiting insights in the past. “You can’t always see the problems when you’re looking at rows and columns”, says Vishniakova. “Cortex is helping us see the source more directly to improve data quality — and that will lead to even faster decision making in the future.”

Plus, better data quality means users can get the same insights with less data processing, reducing runtime and energy consumption. This adds up to make TUI an even more sustainable business.

While it’s hard to tell the exact impacts of these use cases at this stage, the TUI Technology Analytics team is confident the benefits will be extensive for employees and customers. More flexible technology and improved data quality leads to reliable answers from the AI interface. ““That will definitely shorten time to implementation and make us even more responsive to customer needs,” says Vishniakova.

TUI already has two powerful AI use cases in user acceptance testing. However, it has further plans to take advantage of AI tools and open up data opportunities for all. 

Currently, the team is experimenting with applying Cortex to its metadata store and creating a semantic data catalogue layer. This would help teams search all company data and understand what data sits where and who’s responsible for it — without having to go in and access that data directly. For example, this tool could show at a glance where specific application tables sit, where sensitive data can be stored and who can offer access to it. 

“We’re building a web across our data to connect everyone — from interns to our C-level,” says Stefanska. “It all adds up to a platform that can answer questions about data that our previous systems never could. That means our people can be even smarter in how they work with data and serve our customers..”