Expedia headquarters in Bellevue, Washington.
getty
Expedia Group is one of the world’s largest online travel platforms, a 30-year-old marketplace that connects hundreds of millions of travelers with hundreds of thousands of partners across hotels, airlines, cruises and car rentals. The company operates a portfolio of brands including Expedia, Hotels.com and Vrbo, generating approximately $14.7 billion in revenue in 2025. It is, as Chief Technology Officer Ramana Thumu describes it, “a very exciting but very complicated business.”
Thumu joined Expedia Group 16 months ago, bringing more than two decades of experience scaling data-driven platforms at companies including Fanatics and eBay. His remit spans engineering, data, artificial intelligence, cloud platforms and cybersecurity. In a recent conversation, he discussed how Expedia is modernizing its technology foundation, deploying AI across both sides of its marketplace and reimagining what travel could feel like in the years ahead.
Managing a Two-Sided Marketplace
The complexity of Expedia’s business stems partly from the dual nature of the platform. On the traveler side, Thumu’s team applies data assets across the entire shopping funnel, from the homepage through booking and post-transaction service. “We put the right product and the right sort in front of them,” he noted. On the supply side, integrating with hotel, airline, cruise and car rental partners requires solutions tailored to varying levels of technology maturity.
Expedia CTO Ramana Thumu
Expedia
To help partners act on marketplace signals more effectively, Expedia built an intelligence engine called Scout. It surfaces personalized merchandising and pricing recommendations through a simple interface, allowing partners to approve or pause suggestions with minimal effort. “We take all the complexity away from them,” Thumu explained. “It’s all AI-driven, applying all the data assets and the closed-loop demand signals that we have.”
Laying the Foundation for AI
Thumu is emphatic that successful AI adoption requires as much attention to culture and operating model as it does to technology. Expedia has embedded dedicated AI squads within business units, pairing expert AI engineers with business champions who retool processes alongside them. “Change is tough,” he observed. “How do you take the organization through the change while going through this transformation?”
The company has invested heavily in AI literacy across functions and recently brought in external talent, including a hire from Google with deep AI and machine learning experience, to serve as a change agent. More than 90% of Expedia’s employees now interact with AI in some form, from campaign managers to legal and finance teams.
A Data Foundation Built for Scale
Underpinning all of it is a data infrastructure that streams billions of events daily across the marketplace. Thumu’s team is building trusted data assets with clear lineage and discoverability, serving both business analysts and the AI models that power new experiences. “For any AI journey to be successful, it starts with the data at its core as a foundational building block,” he underscored. Above that layer, Expedia is retooling its technology stack to support agentic workflows, incorporating agents, MCPs and skills that can power the next generation of experiences.
Innovation Across the Traveler Journey
On the customer-facing side, AI has already transformed Expedia’s service operations: 80% of contacts are resolved in under 60 seconds, with a 92% first-contact resolution rate across 31 languages in 50 countries. The longer-term vision is a travel companion built on three principles: personalization, prediction and proactivity. “AI allows us to tailor the experience in a way that simply was not possible before,” Thumu remarked. The goal is an orchestrated, agentic experience that anticipates what travelers want before they ask and helps them manage changes mid-trip, all while keeping the traveler in control.
Reinventing How Engineers Work
Internally, 92% of Expedia’s engineers have adopted AI coding assistants, and more than 10% are achieving two to five times productivity improvements. Thumu described a senior engineer at the company’s India Tech Center who built a personal AI workbench of roughly ten agents, supported by 50 to 60 specialized skills, enabling him to deploy features at three to five times his previous velocity. “His job has changed from writing lots of code to orchestrating these agents,” he said.
Scaling that model is now a priority. “Once more and more people do that, there is an inflection point,” Thumu added. “That is when a massive hockey stick of innovation and change management happens.” With more than 200 petabytes of data and a technology stack being purpose-built for agentic experiences, he believes Expedia is well-positioned to reach it.
Peter High is President of Metis Strategy, a business and IT advisory firm. He has written three bestselling books, including his latest Getting to Nimble. He also moderates the Technovation podcast series and speaks at conferences around the world. Follow him on X @PeterAHigh.
