Development cut to a fifth: how Omio remade travel with Codex and ChatGPT!

Hey everyone, it's Shiichan! Today I found a really exciting real-world example of how AI can reshape travel itself.

OpenAI News openai.com

What was announced?

OpenAI's News published a case study on Omio, a leading European multimodal travel platform. Omio lets you search and book journeys across trains, buses, ferries, and flights, connecting more than 3,000 transportation providers across 47 countries.

Omio is using OpenAI models in two directions: one is customer-facing "conversational travel", and the other is speeding up how they build things internally with Codex. They're trying to turn travel from something you "search" into something you "talk through".

Why it matters

Travel search used to mean typing in an origin and a destination, then clicking through filters one by one. Once you mix transport modes, comparing "is the train or the flight faster?" or "what about transfers?" quietly gets painful.

Back in 2023, Omio launched one of the earliest travel experiences available inside ChatGPT. They connected ChatGPT directly to Omio's transportation inventory and booking systems, so a conversation could surface real, bookable journeys. The key part is grounding: the AI isn't guessing, it's anchored to live availability and pricing.

What changes

Travelers can just ask in plain language, like "What's the fastest route from Rome to Florence?" or "Should I take a train or a flight from Paris to Barcelona?" What comes back are real options you can book right there.

That means AI becomes the interface layer between customers and real-world transportation systems. Omio sees this as the doorway to a new category of conversational commerce. Even people who don't travel much can plan a trip as easily as chatting with someone for advice.

Dive Deep

The internal story is fun too. Omio started by rolling out ChatGPT to everyone in the company so people could experiment and learn where it helps. From there they embedded Codex deeply into engineering workflows, and they're increasingly extending it to non-technical functions.

Codex shows up across the whole lifecycle: preliminary research, architecture and planning, coding, automated testing, code reviews, and maintenance.

And the numbers are striking. According to OpenAI's case study, the technical effort required dropped to about 20 percent of before (a fifth). Projects that used to take several engineers a full quarter now land at roughly one engineer in one month.

Still, Omio CTO Tomas Vocetka is clear that people stay at the center:

The responsibility and accountability stay with people. AI helps us develop faster, analyse faster, and make decisions faster, but people stay in charge.

Responsibility and accountability stay with people; AI just helps you build, analyze, and decide faster. (These figures are vendor-reported by OpenAI, so keep that in the back of your mind.)

Wrap-up

• Omio is a travel platform spanning trains, buses, ferries, and flights, covering 3,000+ providers across 47 countries
• For customers, it wires ChatGPT into live inventory and booking to deliver conversational travel with real, bookable itineraries
• Internally, it runs Codex across the whole development lifecycle and is expanding it beyond engineering
• OpenAI's reported figures: technical effort down to about a fifth, and quarter-long projects compressed to one person in one month
• The CTO stresses that people stay in charge of the decisions

If you're thinking about how to weave AI into a travel product, or you want Codex in your team's dev flow, this one paints a concrete picture!