Amadeus President Decius Valmorbida to Address AI’s “Infinite Search” Challenge at Skift Global Forum 2026

Decius Valmorbida, the esteemed President of Travel at Amadeus, is slated to be a pivotal speaker at the Skift Global Forum in New York City, taking place from September 22-24, 2026. His address is anticipated to focus on one of the most pressing and rapidly evolving challenges facing the global travel technology sector: the phenomenon of "infinite search" driven by sophisticated artificial intelligence agents. This issue highlights a growing chasm between traditional travel systems, meticulously designed over decades for human interaction patterns, and the relentless, exhaustive querying capabilities of modern AI. Valmorbida’s insights will be crucial as the industry grapples with the implications of these advanced technologies on infrastructure, operational costs, and the very economics of travel distribution.

The Dawn of AI-Driven Travel Search and Its Disruptive Potential

The integration of artificial intelligence into various facets of the travel industry has been a gradual process, evolving from rudimentary chatbots and personalized recommendation engines to today’s highly advanced generative AI and large language models. Historically, travel search engines and global distribution systems (GDSs) were built with the assumption that human users would conduct searches. A human traveler typically has a specific destination, dates, and perhaps a few preferred airlines or price points in mind, leading to a finite and manageable number of queries. These systems, including Amadeus, Sabre, and Travelport, optimized their architecture to process these human-scale requests efficiently, aggregating vast amounts of real-time flight, hotel, and car rental data from thousands of providers.

However, the advent of powerful AI agents has fundamentally altered this paradigm. Unlike humans, an AI agent does not experience fatigue, cognitive overload, or budget constraints in its initial search phase. When tasked with finding travel options, especially for complex itineraries or price comparisons, these agents can systematically explore every conceivable fare combination, route, and date permutation. This capability, while seemingly beneficial for uncovering optimal deals, poses an unprecedented operational strain on the underlying travel infrastructure.

An illustrative example, recently highlighted by Skift, involved a coding agent unleashed upon Etihad Airways’ website. This single agent, unconstrained by human search heuristics, meticulously trawled the site, ultimately returning an astonishing more than 880,000 distinct fare options for a single trip. Such a volume of queries, if replicated across numerous AI agents simultaneously, could overwhelm even the most robust systems, leading to increased operational costs, potential service disruptions, and data integrity challenges for airlines and GDS providers alike.

Amadeus’s Scale and the Challenge of System Overload

Decius Valmorbida emphasized during a recent conversation with Skift that operating at scale has always been Amadeus’s core competency. The company’s systems are engineered to handle an immense throughput, currently processing approximately 150,000 transactions per second globally. This formidable capacity underscores its critical role in facilitating billions of travel bookings annually. Yet, even systems of this magnitude face an existential threat from the "infinite search" problem.

The gap between travel systems built for human behavior and the machines now using them is not merely an inconvenience; it represents a fundamental architectural incompatibility. Traditional systems are designed to serve millions of individual human requests, each typically leading to a booking or a narrowed search. AI agents, however, can generate millions of exploratory requests, many of which may not lead to an immediate booking but still consume significant computational resources. Each query, regardless of its ultimate outcome, incurs processing power, database lookups, and network bandwidth, translating directly into operational costs for the airlines and GDSs.

The financial implications are substantial. Airlines pay GDSs for each transaction processed, and while the exact cost per query varies, an exponential increase in machine-generated queries could lead to an untenable rise in distribution expenses. Moreover, the sheer volume of data being pulled and processed could necessitate significant investments in server infrastructure, network upgrades, and advanced data management solutions, all to accommodate requests that might not yield a direct revenue stream.

Valmorbida’s Strategic Solution: Precomputation and Predictive Analytics

In response to this emerging threat, Valmorbida has outlined a strategic solution centered on precomputation rather than answering each query live. This approach represents a significant paradigm shift in how travel data is managed and distributed. Instead of performing a real-time database lookup for every single request, Amadeus aims to leverage its vast data processing capabilities to precompute and cache a large set of possible fare combinations, routes, and availability options.

This precomputation strategy would involve:

1. Proactive Data Aggregation: Continuously collecting and analyzing flight, hotel, and other travel data from thousands of providers.
2. Predictive Modeling: Employing advanced algorithms to anticipate common search patterns, popular routes, and potential price fluctuations. This allows for the proactive generation of search results before a query is even made.
3. Dynamic Caching: Storing these precomputed results in highly optimized, low-latency caches. When an AI agent (or even a human user) initiates a search, the system can first check the cache for relevant results, drastically reducing the need for real-time, resource-intensive database queries.
4. Intelligent Indexing: Organizing the precomputed data in a manner that allows for rapid retrieval based on various parameters such as origin, destination, dates, airline, and price range.

The benefits of precomputation are multifaceted. For Amadeus and its partners, it promises enhanced system efficiency, reduced server load, and significantly lower operational costs associated with query processing. For airlines, it means a more stable and predictable distribution environment, less vulnerability to "denial-of-service" style attacks from exhaustive AI agents, and potentially faster response times for their legitimate customers. For Online Travel Agencies (OTAs) and other travel sellers utilizing Amadeus’s platform, it could translate into faster search results and a more seamless user experience, even for complex queries.

Of course, implementing such a sophisticated precomputation system is not without its challenges. Maintaining data freshness and accuracy is paramount in the volatile travel industry, where prices and availability can change by the minute. Amadeus would need to develop robust mechanisms for real-time updates to its cached data, ensuring that precomputed results remain relevant. The storage requirements for such a vast dataset would also be immense, necessitating significant investments in cloud infrastructure and data management technologies. Furthermore, the complexity of the algorithms required to intelligently precompute and dynamically update results would push the boundaries of current travel tech capabilities.

The Skift Global Forum 2026: A Confluence of Industry Leaders

The Skift Global Forum, where Valmorbida is set to speak, is widely recognized as the premier annual executive conference in the global travel industry. Held annually in New York City, it convenes an exclusive audience of CEOs, founders, and innovators from across the travel ecosystem, including airlines, hotels, tour operators, technology providers, and investors. The forum is renowned for its forward-thinking agenda, tackling the most pressing issues and emerging trends shaping the future of travel.

In 2026, the discussion around AI’s transformative impact, particularly on core infrastructure like search and distribution, will undoubtedly take center stage. Valmorbida’s presentation is expected to be a cornerstone of these discussions, offering a glimpse into how one of the industry’s giants is proactively addressing a challenge that affects every player. The forum provides an invaluable platform for industry leaders to share insights, forge partnerships, and collectively strategize on navigating the complexities of technological advancements. Past forums have explored themes such as sustainable travel, personalized experiences, the metaverse in travel, and the evolving role of human connection in an increasingly digital world. The "infinite search" problem perfectly aligns with the forum’s tradition of addressing foundational shifts in the industry’s operational backbone.

Broader Implications and the Future of Travel Distribution

The implications of AI-driven "infinite search" and Amadeus’s precomputation solution extend far beyond mere technical efficiency.

For Airlines: This shift could fundamentally alter revenue management and dynamic pricing strategies. If AI agents can instantaneously discover and exploit minute price discrepancies across vast datasets, airlines will need even more sophisticated algorithms to set and adjust fares in real-time, potentially leading to more fluid and competitive pricing. The ability to manage query load will also reduce costs associated with GDS fees and server infrastructure.

For Online Travel Agencies (OTAs) and Metasearch Engines: These entities heavily rely on GDSs and direct airline connections to power their own search capabilities. If GDSs like Amadeus can provide precomputed, optimized results, OTAs could offer faster and potentially more comprehensive search experiences to their users, enhancing their competitive edge. However, it also means that the "secret sauce" of finding the absolute best deal might shift from the OTA’s proprietary algorithms to the underlying GDS’s precomputation engine.

For Global Distribution Systems (GDSs): Amadeus’s proactive stance highlights the critical need for GDSs to evolve from mere data conduits to intelligent data processors. Their role will increasingly involve not just aggregating real-time data but also enriching it, analyzing it, and presenting it in optimized formats, whether for human users or AI agents. This necessitates significant R&D investment and a shift in business models. Competitors like Sabre and Travelport are undoubtedly pursuing similar innovations to stay competitive.

For Consumers: In the long run, these advancements could lead to more efficient and comprehensive travel search experiences. While direct AI agent interaction for booking is still nascent, the underlying improvements in data processing could result in quicker loading times, more accurate results, and potentially more competitive pricing as the system becomes more optimized. However, there’s also a risk of increased opacity if precomputed results become the norm, potentially obscuring the full range of options or the exact methodology behind a recommendation.

Ethical Considerations and Regulatory Oversight: As AI plays a larger role in determining travel options, ethical considerations surrounding algorithmic bias become paramount. Could certain precomputation models inadvertently favor specific airlines or routes, or could they perpetuate existing biases in historical data? Ensuring fairness, transparency, and consumer protection in AI-driven travel search will be a growing concern for regulators.

The "infinite search" challenge posed by AI agents represents a crucible moment for the travel technology industry. Decius Valmorbida’s upcoming address at the Skift Global Forum 2026 is expected to not only shed light on Amadeus’s strategic response but also spark a broader industry dialogue on how to harness the immense power of AI while safeguarding the stability, efficiency, and fairness of global travel distribution systems. The future of travel search will not be about answering every query live, but about intelligently anticipating and precomputing the vast universe of possibilities.