OAG sits at the heart of the global aviation ecosystem. Every major airline, airport, online travel agency (OTA), global distribution system (GDS) and many emerging AI-driven travel systems depend on OAG schedules, status, fare and capacity data to make decisions.
As the industry shifts from reporting on what happened to predicting and acting on what happens next, the demands on OAG's data foundation have changed fundamentally.
Legacy architecture built for batch delivery of reference data could not provide the real-time context, governed access and machine-learning-ready pipelines required to power predictive models and emerging agentic travel systems.
A planned four-year migration timeline risked leaving OAG and the customers who built it a step behind the market.
The goal was to modernize the foundation, accelerate predictive intelligence capabilities, and position OAG as the trusted data and intelligence layer for an industry moving toward autonomous decision making.
The approach:
OAG and Thoughtworks executed a strategic platform transformation, consolidating fragmented data ecosystems into a unified, governed foundation on Databricks with Unity Catalog at its core.
To power this transformation, Thoughtworks recommended deploying Databricks on Azure. This leveraged OAG's existing Microsoft ecosystem while utilizing Apache Spark for high-performance processing. By removing low-level infrastructure management, Databricks provided an expandable ecosystem built to scale. The work spanned three reinforcing layers:
- A modern data platform engineered for high-velocity aviation feeds.
- A governed, machine-learning-ready data product layer covering schedules, status and fares.
- An intelligence layer that turns that data into predictive signals customers can build into their own operations.
One technical proof point is live status prediction that refines estimated departure and arrival times in real time, using OAG's global status feed combined with other data sources. For airlines, that means earlier and more accurate visibility into delays, better turn planning and fewer downstream disruptions. For the platforms, OTAs and agents that depend on OAG, it means richer, predictive intelligence delivered through the same trusted data layer they already consume.
The outcome:
Reduced time to onboard new data feeds by more than 50%,
accelerating the pace at which OAG can improve data quality for
customers and stand up more refined intelligence products.
Compressed a four-year legacy migration into a 12-month delivery window without disruption to mission-critical customer feeds.
Refined estimated time of departure (ETD) and estimated time of arrival (ETA) in real time with predictive status intelligence to give airlines and travel platforms earlier, more accurate operational signals.
Decreased issue remediation from five steps to two through unified governance, automated quality dashboards and end-to-end tracing across the platform. This improves our responsiveness to customer queries.
OAG data sits inside the decisions airlines, airports and travel platforms make every minute of every day. With Thoughtworks, we've modernized the foundation so we can do more than tell our customers what happened. We can tell them what's about to happen, and give them the intelligence to act on it. A governed, real-time data platform is the prerequisite for predictive and agentic aviation. The foundation is built. Live predictions are in production. The intelligence layer that will power the next generation of airline, airport and travel operations is being built on top of it.
