LATAM Airlines

300+ people, 50 designers, no existing design system. A single product vision that had to survive that level of complexity without fragmenting.

LATAM Airlines was rebuilding its entire digital experience from the ground up, a platform serving millions of passengers across Latin America, in multiple languages, across web, mobile, and app simultaneously. The scale was unusual even by enterprise standards: 300+ people, 50 designers, a single product vision that had to survive that level of organizational complexity without fragmenting. The risk in projects of this size isn't technical. It's coherence.

Within the broader program, I owned a critical cluster of the experience: the marketing homepage, login, sign-up, user profile, and the entire Contact Center and help desk ecosystem. These weren't peripheral surfaces. The homepage was the primary commercial entry point for the platform, and the Contact Center was the last line of defense before a passenger escalated to human support.

I led a design duo within my verticals, a deliberate pairing model that allowed for faster iteration, tighter quality control, and continuous design critique without the overhead of a larger team structure. In a program of 50 designers, the ability to move with focus and precision inside your own scope while staying coherent with the whole was a discipline in itself.

One of the defining challenges of the project was that there was no existing Design System to inherit. We built it from scratch, simultaneously designing the product and establishing the component library, patterns, and documentation that would serve 50 designers across the entire program. Every component I contributed had to be abstracted, stress-tested, and viable for reuse across contexts I couldn't fully anticipate. That forced a systems thinking approach to every individual screen, not just the ones I owned.

The most consequential decision I made on this project wasn't a design decision. It was a methodology decision. Every screen I designed was built with a full accessibility storytelling layer: a structured narrative defining how each element would be read by a screen reader, in sequence, for a blind user. Treating it as a compliance checkbox would have been easier. Instead I used it as a design constraint. If a screen couldn't be explained linearly without confusion, it wasn't clear enough visually either.