The World’s Airlines Rely on Graphs to Offer Better Ticket Prices and Travel Options for Customers

The Challenge

ATPCO is the world’s leading provider of pricing and retailing content, tools and services to airlines, global distribution systems, travel agencies and technology companies. By using compelling, rich content, its retailing solutions are also setting the standard by which airlines differentiate their products and channel partners create informative, next-generation shopping displays.

To deliver such services means dealing with immense amounts of intricate information – information that represents huge value for its users. ATPCO’s Chief Architect, Navid Abbassi said, “Getting a competitive price for a plane journey involves a large amount of complex data processing that the traveler just never sees. So it’s a big job, and an important one. Just one of our many airline customers estimates that, every day, if you add up all the various markets, flights, seats, and options in what they put in the sky, it’s over 100 billion product permutations.”

That’s why it’s so incredibly important that ATPCO do everything possible to always deliver the very best fare and offer management tools to meet the ever-changing needs of their customers.

The Solution

To achieve maximum transparency and functionality, the team knew they needed to interconnect many siloed data sources and applications via a core pricing engine.

ATPCO’s internal R&D team needed a powerful, robust technology to help its customers better utilize their vast data resources. Finding the perfect basis for such a capable pricing engine was actually something of a lucky accident.

David Peart, Enterprise Innovation Architect at ATPCO, recalls evaluating NoSQL databases back in 2010, when he realized that Neo4j was a natural fit to provide the data architecture, power, capacity and scale to represent the millions of data relationships they need to model and understand.

“A trip from Newcastle, UK to Lexington, Kentucky involves multiple possible routes and prices,” Peart said. “But I was able to use graphs to represent different airports as nodes and all the scheduled flights as relationships, allowing me to complete this very complex task far more easily than with other tools.”

That early route planning experiment started a journey that’s led to a special graph-powered pricing engine. Abbassi added, “We now use Neo4j as the core of at least five of the main data services we now offer, from fare management to air travel tax calculation.”

Download Case Study