The cabin at altitude

Delta had spent years getting the fundamentals right. Free wi-fi when competitors charged for it. A loyalty program travelers could actually understand. The airline that felt like it was genuinely trying. None of that ambition had been asked of the in-flight entertainment system. It had been stable. It had done the job. But stability isn’t progress. The experience was a decade behind the partners it was meant to serve.

The ask was to close that gap: bring the IFE to the altitude the partnerships deserved and give those partners proper homes. My assumption going in was that the experience challenge was the main event. I was partly wrong about that.

The hardest design problem wasn’t the interface. It was AI.

Not whether to include it (that decision had been made above us) but how to limit it to something a passenger might actually need. Every AI integration in consumer products at that point was either a novelty or a liability. The question was what a traveler genuinely needs AI to do at 37,000 feet.

Two ends of the spectrum defined the answer.

In the first, the AI earns its place by knowing what you care about and surfacing it before you think to ask. In the second, by guiding you through rebooking without making you fight a system that doesn’t care whether you make it home.

Proactive

Before you think to ask

Reactive

When something breaks

Delta’s sweet spot

Scoped to the journey, not the product

Both required continuity, not just within the IFE screen, but across your phone, the cabin system, and the cloud. The cabin needed to feel like one coherent thing, aware of where you were in your journey, able to hand off between surfaces without starting over. That’s what “connected” meant. Not a feature. A posture the whole experience had to hold.

This was a vision project, not a spec. Its job was to make a room of decision-makers believe before anyone committed to building. That changed what every subsequent decision was actually for.

We scoped the AI to the journey, not the product. Flight-relevant needs only: no general assistant, no off-topic queries. The constraint made it more useful, not less. An AI that knows you’re on a plane and knows where you’re going can give you answers a general-purpose assistant can’t. The specificity was the point.

The visual direction had to do the same work for the content partners. If these platforms were going to have homes in this experience, those homes needed proper architectural presence, not catalogue entries buried in a menu. The experience had to feel like it was built at the same level of ambition as what it was hosting.

Delta’s CEO announced the connected cabin vision at CES in January 2025. That’s the headline.

Delta now has 18 months of commitments made from a CES stage, and the market has moved in ways that raise the stakes. Competitors have closed the free satellite wi-fi gap on international flights. The last friction point where Delta had a clear lead has been matched. The roadmap our vision helped create has to deliver against more pressure than existed when we started.

If I ran this project again, I’d partner differently with the Delta strategy lead. I spent too much energy trying to guide the product direction when my real value was as a sparring partner and execution arm, someone who could make her thinking undeniable rather than someone steering it. The guidance I was offering was adding friction to the wrong side of the work.

What this project changed for me: on vision work, the measure isn’t whether the experience is refined enough to ship. It’s whether the vision is clear enough to believe. Business momentum and internal conviction matter more than interaction fidelity when the output is a funding decision. I came in thinking like a product designer. I should have been thinking like someone trying to build conviction in a room, and trusted that the craft would follow once the conviction was there.