Holiday Extras reports that using Codex to rebuild its main mobile app produced faster development cycles and improved booking performance. The company says the project reduced a rebuild timeline that once took roughly six months to a working version completed in two weeks, and shifted the release cadence from annual major updates to weekly releases.
App rebuild and development speed
According to Holiday Extras, the rebuilt app delivered a 32.5% increase in open-to-book conversion and lowered the error rate from 19% to 9%. The team attributes these changes to a new development approach supported by Codex: small, documented packages for reusable foundations while engineers focused on architecture, APIs, security and workflows.
Work on iOS and Android features was carried out together to avoid one platform lagging behind. Liam Thompson, VP of AI Innovation, said that Codex changed how product ideas were translated into software: “It’s not a question of can I build it anymore. It’s how I would build it.”
Operational effects and company access to AI
Holiday Extras established company-wide access to ChatGPT Enterprise and used Codex to extend capabilities into software development and internal systems. Leadership encouraged experimentation with guardrails: engineers created the technical controls that made broader use practical, while staff across functions could apply operational knowledge directly to tools.
The company describes concrete examples where faster development improved customer outcomes. One customer service fix was prepared, approved and submitted in days after an agent identified a missing list view; the customer changed a one-star review to five stars. David Lee, Director of Innovation and Growth, described an automated system that turns airport-transfer chats into actionable tickets to surface supplier issues.
CEO Matthew Pack framed the shift as organisational: “This isn’t about one or two people anymore. It’s about taking the entire team on the journey.” Holiday Extras says it plans to expand AI-administered systems and is asking leadership to reassess the scale of opportunities the company will pursue.
Original source: OpenAI News