How Virgin Atlantic Ships Software Faster with Codex
Virgin Atlantic uses Codex to strengthen test coverage, accelerate refactoring, and ship customer-facing software with greater confidence.

Company size: Enterprise
Region: Europe and the UK
Industry: Travel
Product: Codex
78–80%
Reduction in codebase size during legacy system refactoring
Nearly 100%
Unit test coverage for the new app
30
Minutes to refactor legacy codebase, down from 2 weeks
Virgin Atlantic used Codex to launch a brand-new mobile app in time for the Christmas travel rush — one of the highest-risk periods of the year for software bugs.
“We’re an operational airline, so we have to be really careful about the software we release to customers,” explained Neil Letchford, Vice President of Digital Engineering at Virgin Atlantic. “People are using it to get on the plane. They need to be able to check in, and they need to be able to board.” With Codex’s help, the team achieved near-complete unit test coverage within the launch window and shipped with zero P1 defects.
Codex is also helping the team refactor years of legacy code at unprecedented speed, while analytics teams across the airline are building tools directly on top of the company’s data warehouse.
00:00
“Using Codex to improve app quality before it gets into customers’ hands has been a game changer for us.”
—Neil Letchford, Vice President of Digital Engineering
Getting the launch over the line
Virgin Atlantic launched the new mobile app in beta during the Christmas period and moved it into production a few weeks later. “The quality level was exceptional because of the test coverage,” Letchford said.
That level of quality is hard to achieve under deadline pressure. Engineering teams working to fixed launch dates often have to cut scope or compromise on testing just to ship on time. Codex helped Virgin Atlantic do the opposite: the app went live with zero P1 tickets.
“From an enterprise point of view, those conversations are really interesting. We’re talking to senior leadership, trying to tell everyone that things are good to go, all green lights,” Letchford said. “These are all new things we’re not quite used to doing. With Codex, things don’t get held up.”
Beyond mobile
The same improvements are showing up in legacy code as well. Codebases that the team had maintained for years can now be refactored in hours instead of weeks. “In some cases, we’ve seen 78% to 80% reductions in the size of codebases with Codex,” Letchford said.
The speed is now beginning to outpace the rest of the delivery process. In a recent sprint, one of the team’s frontend leads — a heavy Codex user — built a fully functional frontend app in a week from a Figma prototype, with placeholder backend logic. “The Scrum Master complained that we didn’t even have the backend tickets ready to start work on them,” Letchford said.
That efficiency is extending into data work as well. “Codex has really helped us unlock and de-risk moving disparate databases into our core data warehouse,” said Richard Masters, Vice President of Data and AI. With Codex, analytics teams can now rapidly prototype internal apps directly around the company’s data warehouse.
“You can prototype data in just a few hours, or even in a workshop,” Masters said. Teams across network planning, customer experience, and engineering and maintenance are now using Codex to build their own internal applications, instead of sending requests to the central Data and AI team.
“The efficiency gains we’re seeing in refactoring are huge: work that used to take two weeks can now take 30 minutes to an hour.”
—Neil Letchford, Vice President of Digital Engineering
What’s next
Codex is reshaping how Virgin Atlantic ships software, from the apps passengers use at the gate to the data platforms behind them. For Letchford, the next question is how to manage a pace of engineering that starts to outstrip the rest of the delivery process.
“How do we start to scale this beyond just one pocket and across the entire software development lifecycle?” he said. “It’s a really interesting question — and a good problem to have.”
“The trajectory of Codex is no longer just for pure engineers. It’s becoming a true tool for everyone.”
—Richard Masters, Vice President of Data and AI