6 minutes read
What this article talks about
- Why platforms, not AI models, determine outcomes by controlling access, orchestration, reliability and what can actually be booked.
- How modern infrastructure removes industry complexity through unified data, normalization, operational shielding and scalable NDC integration.
- How AI acts as an enhancement layer that depends on strong platforms for data quality, personalization and reliable automation.
Underlying infrastructure will determine who wins in an AI-enabled travel industry, says Tracey Hempstock, Director, NDC Offer & Order at Travelport.
As we’ve already seen in this blog series, AI is great for recommendations and travel planning, and for automating low-risk operational tasks. But platforms are the ultimate differentiators, shielding sellers and travelers from complexity, controlling what can actually be booked, and making AI possible in the first place.
Social media influences and inspires. AI enables discovery, detailed itineraries, and multiple options. But it’s the platform and infrastructure that deliver commercial reality—determining access, bias, and commercial leverage. Travelport is, first and foremost, a platform business. The platform is the core enabler from which everything flows.
Access, orchestration, and safety: ‘air traffic control’
How do platforms influence commercial outcomes? Think of a platform like air traffic control, directing which flights (content) can enter the airspace and eventually land. Where and when they appear, and how safely and smoothly everything moves into place. Even if airlines and other travel providers have strong offers, it’s the platform that enables access, flow and sequencing—which ultimately determines what travel agencies and travelers see first.
A simple retail metaphor: platforms are like supermarket shelves, with the right products (based on whichever criteria you use) placed at the consumer’s eye level, catching their interest and, potentially, selling as a result. Visibility can matter as much as product quality. Put another way: AI amplifies interest and discovery, whetting the appetite, while the platform prepares the meal using the freshest ingredients, serves it, presents the bill, and takes payment.
The foundational layer: making sense of complexity
For our customers, and by extension the travelers they support, Travelport’s value lies in our platform and infrastructure; with the access, orchestration, reliability, and ease of use that we provide. Investments and developments over the last three years mean our platform enables scalable, AI-ready travel distribution through unified data access and processing.
This “foundational layer” aggregates, normalizes, and standardizes fragmented NDC and other content. This removes the need for customers to manage multiple direct integrations, different schema versions, and multiple external IT providers. It’s an incredibly complex environment, and our platform removes that complexity. It provides “operational shielding”—insulating customers from industry fragmentation, API differences, schema upgrades, internal IT teams and third-party tech providers, ongoing airline changes, and integration maintenance.
Direct airline integration is complex. This fragmentation across multiple APIs, tech providers and schema variations makes a DIY approach inefficient and difficult to sustain at scale. By contrast, the Travelport NDC platform ensures data quality, consistency, and completeness thanks to strict implementation standards including 120+ NDC capability checks. This leads to a more reliable and uniform customer experience.
Importantly, our platform enables critical functionality at scale, dealing with issues like schedule changes and handling flight disruption, across all integrated carriers. That’s something other providers can only support partially. Our infrastructure also supports fairness, and unbiased content display, governed by airline agreements. Results are presented based on objective criteria, such as price filters, rather than commercial bias.
Travelport’s architecture is modern and forward-looking, optimized to ensure reliability and scalability. And why is all this so important? Those characteristics form the foundations needed to enable AI to perform at its fullest potential.
THE CUSTOMER VIEW: UNIFIFI
UNIFIFI connects travel agencies with global flight content from airlines and GDS systems. It uses AI as a supporting tool rather than a decision maker, enhancing processes without replacing platform control or core distribution logic. “We have a user-centric distribution model, tailoring content based on customer type, such as corporates versus OTAs, and specific needs such as flexibility, pricing and ancillary services,” says Afra Wang, Co-founder and COO. UNIFIFI uses historical data and behavioral insights to filter and prioritize relevant content, improving speed, conversion rates, and the overall user experience. With content flow orchestrated and filtered dynamically, the right offers can reach the right audience, rather than all options being presented equally.
AI can enhance automation and performance, but only works effectively when built on strong, modern infrastructure. Weak or legacy systems limit AI performance.
Afra Wang, Co-founder and COO at UNIFIFI
She says AI can enhance the infrastructure by proactively identifying and resolving issues, to help maintain system stability, API performance, and reliability. But the platform remains in control. “AI is an enhancer, not a decision-maker. AI improves efficiency and provides new insights, supporting fairness and transparency, for example, by enabling unbiased ranking and clearer distribution rules. But it lacks full real-world context and relies on underlying systems and data to function properly. Platforms still govern content access, standards, and real-time data—AI depends on them, rather than replacing them.”
The enhancement layer: AI
As Afra Wang of UNIFIFI indicated, AI can best be seen as an “enhancement layer” to complement and extend the capabilities of your core platform. It’s a way to surface new commercial opportunities, create more personalized offers, and improve efficiency (from smarter querying via well-designed chatbots, to enabling NDC schema upgrades). But we should always remember that AI is wholly dependent on the underlying platform’s data integrity and structure. In UNIFIFI’s case, with the business focused on user-centric content orchestration, it’s the platform that determines which content is shown based on customer type, behavior, and needs, using data-driven filtering to deliver the most relevant options.
That’s what it boils down to: a platform fit for purpose today, while creating leverage for future AI capabilities through the provision of cleaner, better structured, and trusted data—essential for accurate AI-driven personalization, automation, and decision-making.
We’re seeing incredible AI advancements right now. But building your AI capability on weak foundations can only end in failure or, at the very least, wasted time and money. What matters is the underlying platform, the core GDS, and data integrity enabling integration and orchestration. All future services—and the commercial outcomes they enable—will depend on fast, accurate, and reliable underlying travel data.
This is the third in a blog series exploring how AI is transforming travel. The next blog describes the concept of the “amplified agent”—why human judgment + AI = competitive advantage.