ProVida App.
A pension app has the hardest user base in fintech: it has to work for a 20-year-old opening their first account and a 70-year-old managing their retirement — on the same screen. This is how I rebuilt ProVida from scratch for MetLife, and helped it reach 1M downloads.
A retirement product its own users didn't trust.
ProVida is one of Chile's pension fund administrators, managing retirement savings for a huge cross-section of the country. For most of its users, the app was the only window into the single most important number in their financial life — and that window was failing them.
The data was unambiguous: NPS was very low. Users were frustrated with the old version of the app and felt limited by it — they couldn't see their savings clearly, and they couldn't manage their money the way they needed to. The brief from MetLife wasn't a refresh. It was a complete rebuild, from scratch, with new features.
Two problems, one screen.
The redesign had to solve two problems at the same time:
- A trust problem. Users needed real transparency into their retirement savings, and the ability to actually administrate their money — not just look at it.
- An accessibility problem. ProVida's users span almost every adult age. Every one of them — regardless of age or financial literacy — had to be able to complete critical financial tasks without help.
That second constraint is the one that shaped everything. It's not hard to design a beautiful pension app. It's hard to design one that works equally well for someone opening their first account and someone drawing down their retirement.
The title said Visual Designer. The scope was end-to-end.
I owned every flow and every screen in the product. I presented the work directly to the client, coordinated with the development team based in India, and tested designs with real users. The team was distributed across the world, working in an agile environment — so the design couldn't just be right, it had to be communicated with zero ambiguity.
ProVida is a financial platform serving multiple types of users, each with different features and scenarios tailored to them. To keep errors low and keep a globally distributed team building the same product, I created a map of scenarios by user type — an explicit record of who sees which features, in which states, and why.
"On a product serving many user types, a shared map of who-sees-what is the cheapest insurance you can buy against shipping the wrong thing."
Designing for every age.
The core design problem was range. A 22-year-old and a 68-year-old had to be equally able to check their savings and move their money. I worked the problem on four fronts:
Research with real users
I tested with real users across the full age spectrum — not just the easy-to-recruit ones — so the design was validated against the people who'd actually struggle, not an idealized average user.
Clear, simple visual hierarchy
Legibility and obvious navigation over cleverness. Every screen was built so the primary action was unmistakable, reducing the cognitive load for users who don't navigate apps fluently.
Language without financial jargon
I translated complex pension and financial concepts into plain language any user could read and act on. In a product about retirement savings, the words carry as much of the UX as the layout.
Iterative task testing
I validated — repeatedly — that users of any age could complete the key tasks, and reworked whatever failed. The bar wasn't "looks good." It was "everyone can finish."
A rebuilt product, in real hands.
ProVida is a living product — it keeps evolving to serve its users better. But after a year of work rebuilding it from the ground up, the redesign was in production and in real hands:
For a product that started from a low-NPS baseline, getting a redesigned-from-scratch app into a million hands — and holding a solid store rating across an unusually wide user base — is the outcome that mattered.
What I carry forward.
What worked
Treating "wide age range" as the central design constraint, not a footnote. When you design for the user who struggles most, you almost always build something better for everyone — the clarity that helps a 70-year-old helps a distracted 25-year-old too.
What I'd carry into any product
The scenario map. On any product serving multiple user types — which is most of fintech — an explicit, shared map of who-sees-what keeps a distributed team aligned and keeps errors out of the build. It's the artifact I'd reach for first on a project like this again.
What I'd add with more time
Tighter post-launch instrumentation tied back to the original NPS problem — tracking task completion by age cohort, so the team could see exactly where the accessibility gains held and where the next iteration needed to go.