Holiday Extras, a mid-market travel company serving customers in Europe and the UK, used OpenAI's Codex coding tool to rebuild its main mobile app, according to a customer story OpenAI published on 12 November. The company says the rebuild reached a fully working state in two weeks, progress that previously would have taken more than six months. It reports that open-to-book conversion on the main app rose 32.5% after the rebuild and that the app's error rate fell from 19% to 9%. The project followed company-wide ChatGPT Enterprise access, with Codex extended to building software for customer-facing apps and internal systems. Liam Thompson, VP of AI Innovation and formerly head of design, worked with engineers to turn his understanding of customers into software. Small, documented packages created reusable foundations, while engineers focused on architecture, APIs, security and development workflows. Codex also allowed iOS and Android features to be built together. The main app moved from an annual major release to weekly releases in the first month after the rebuild. CEO Matthew Pack, who builds with Codex himself, says the aim is to give staff who understand customers and operations the means to improve them directly. The figures come from a vendor-published case study and have not been independently verified. The article notes that the conversion and error-rate comparisons span different periods, and it does not disclose the cost of the deployment or the methodology behind the results. Pack said he wants more of the company's systems to be administered through AI, with employees directing the work. For Australian businesses and professionals, the story changes nothing immediately, since Holiday Extras operates in the UK and Europe and its regulatory context is not Australian. The practical lesson is narrower: AI-assisted development worked for this company because of reusable components, security controls and clear ownership. Any Australian business considering similar tools for customer-facing apps should first test a small internal workflow, and should check its obligations under the Privacy Act 1988 and the Australian Privacy Principles before sending customer personal information to third-party AI services.