Nubank

MMXXV–MMXXVI

The Nubank Home screen, with the total balance and the credit card balance

I led design for Nubank’s US launch

May 2025 to April 2026. Nu launched it in the US on September 10, 2026.

Nubank’s US app started as a concept deck that nobody had built. I turned it into a complete product on one Home that stays put, and every Nu market adopted that Home as its core.

11months from a concept deck to a complete product
3states for every widget on Home
1message at a time asking for your attention
EveryNu market adopted the US Home
Role
Staff Product Designer, design lead for the US launch.
Team
Design leads for four business units, and leads in Brazil, Mexico and Colombia.
Owned
Home and its widgets, navigation, profile, settings and cross-sell.
Outcome
A complete US product. Its Home became the core for every market.

The start

A deck the CEO believed in, and nothing built

Nu was getting ready to enter the US and to redesign its app at the same time. The new patterns existed only as a company-wide concept deck: no prototypes, no screens, no flows.

Four business units were building toward the US in parallel, each with its own design lead and its own priorities. Brazil, with most of Nu’s customers, had the company’s attention. Without someone speaking for the US, it would have become a patchwork of other markets’ decisions.

Early testing showed what that meant for customers. New US customers, expats among them, didn’t understand the single account model, were confused by widgets that kept changing, and couldn’t see what to do next.

Two things were fixed from the start. Home is put together by the server, which limits what a screen can do, and the US launch date did not move.

Who was involved

Me
Design lead for the US launch. I owned what connects every product: Home and its widgets, navigation, profile, settings, Journeys, cross-sell and notifications.
Four design leads
One each for the account and savings, the credit card, card management and remittances.
Other markets
Design leads in Brazil, Mexico, Colombia and Nu Global.
The redesign team
They had defined the new patterns, as concepts.
The design system team
My closest partners. Together we built the header, the widgets and their interactions into the components every market shares.
Leadership
The Chief Design Officer, the US Director of Product, the US CEO and the CEO.
Around the launch
Product, engineering, data, risk, legal, finance, marketing, customer experience, operations, people and culture, and business analysis.

26product dashboards

185feature flows

What I inherited across Nu’s markets, with no shared hierarchy. That is the whole company’s product, not a count of what I designed.

A wall of small app screens, hundreds of them in uneven rows

The process

How it unfolded

  1. Month one

    I gathered every design lead the US depended on, and the design system team, to agree on patterns that would connect globally but be led through the US.

  2. One model

    I brought each business unit into one US experience. As the only US citizen in the group, I brought how banking, law and expectations differ here.

  3. One Home

    With the redesign team I turned their concepts into a working Home, with the US as the proof the patterns would hold everywhere.

  4. A new design chief

    When the Chief Design Officer joined, I briefed him on what the US needed. He helped put the US ahead of Brazil for the design system and redesign teams.

  5. Marketing

    I stood up the US marketing team and led it, then brought in a dedicated lead when it was too much alongside the product.

  6. Approval

    After months of pitching other markets and business units for resources, we got global approval and could move fast.

  7. Walk the Product

    Before launch, everyone responsible walked through the live app together and agreed what to change.

  8. April 2026

    The product was complete. Nu launched it publicly on September 10, 2026.

Decision 1

A Home that stays put

The idea I inherited was a smart Home that rearranges itself around what you’re doing. We tested it with people in Brazil, then in the US. Try both.

Widgets stay where people put them. Each one has three states: inactive, active and cross-sell.

GabrielaTotal balance$1,957.35Credit card balance$1,350.75Earn 4.50% APYBoosted SavingsEarn 1.5%CashbackSecure shoppingMy cardsFast and no-feeSend abroad

People thought we were going to take their widgets away.What testing showed, in Brazil and then in the US

A smart widget showing stacked cards and a credit limitA smart widget showing an offer for phonesA smart widget showing cashback as a bar chartA smart widget showing a card and a balance

Before: the smart widgets we tested. Each one looked and behaved its own way.

After: one widget, in three forms.

The Home screen with four numbered parts: the profile at the top, the hero callout, a row of three actions, and the balance widgets

Making it buildable

Every part of Home has one job

A Home that every team plugs into the same way also held up across business units.

  1. 1The profile switch: who you are and which account.
  2. 2The hero callout: the one message that needs you now.
  3. 3The action row: moving money.
  4. 4The content area: your products, as widgets.

Widgets follow two layouts, 2x2 and 4xN, with set spacing and sizes, so a team could build one without asking. Movement stays only where people expect it: virtual cards, and sending money again to the same people, like family back home.

We didn’t over-index on getting it perfect. We simplified so teams could move.What I told the teams at the time
A drawing of the Home grid with its spacing marked, beside three widgets with their measurements
The rules: one grid, set spacing, and the same widget in three forms.
Two rows of widgets built from those rules: balances, savings, cards, transfers and invites
Widgets built from those rules, across products and markets.

Decision 2

Lead the global patterns through the US

Rather than adapt Brazil’s app, I made the US the first product built entirely on the new patterns and the newest design system components.

That meant winning over far more than the US team. It had to work for the CEO and the US CEO, and for the Brazilian product teams, who serve millions of customers with many more products. So Home had to give every business unit and every product a place of its own.

What I led in the US was adopted as the core for every market, each rolling out on its own timeline:

  • Header
  • Widgets
  • Home structure
  • Profile
  • Settings
  • Journeys
  • Cross-sell
  • Legal disclaimers
  • Global actions
  1. United StatesBuilt here first
  2. BrazilAdopted as the core
  3. MexicoAdopted as the core
  4. ColombiaAdopted as the core

Before and after

From scattered flows to one Home

Before, teams had solved the same problems in different ways. After, every product, action and message has a set place.

The earlier Home in Brazil: a greeting, a hidden balance, a row of round actions and stacked rows below
BeforeThe app as it was, here in Brazil.
The US Home: one message at the top, three actions, then the balances
AfterThe US Home.

Decision 3

One message at a time

Taking the movement out of the widgets meant giving it one deliberate place.

Before, offers were scattered and showed up differently across the app. Next Best Action became the single entry point. Messages are ranked by urgency and impact, and only one is shown: the thing that needs the customer now. The next ones wait their turn.

I kept it apart from notifications and from offers, so it never became a copy of either. An offer appears there only when it is the most important thing for that customer. Teams got a component and the rules for choosing the message.

Priority also decides where a message sits: what is urgent goes to the top of Home, and offers go to the bottom.

The callout: The single most important actionable message in the app, with an Action button and Not now
A Home screen from top to bottom: an alert, the actions, two balance widgets, then offers
  1. Urgent
  2. High
  3. Medium
  4. Low

Working together

Walk the Product

With the Chief Design Officer I ran sessions where everyone responsible for the launch went through the product together.

Executives, product managers, designers and engineers saw how the product looked and behaved with the components we had. I recorded the live app, slowed the animations down and went through them frame by frame, but most of it came down to live demos.

The sessions kept growing, because more people wanted to see the changes that might affect their own team and dates. The launch depended on every one of them lining up.

  1. Walk

    Go through the live app, screen by screen.

  2. Decide

    Agree what needs to change.

  3. Name an owner

    One person for each change.

  4. Set a date

    A timeline every team can plan around.

The app

Five screens from the US app

Welcome, Home, For you, Profile and Settings.

The welcome screen, with a picture of a globe above the words Welcome to the Nu way of banking
Welcome
The Home screen, with a first-deposit message, three actions, the total balance and the credit card balance
Home
Further down Home: an offer, then five widgets for savings, cashback, cards, sending abroad and invites
For you
The profile screen, with a country switch at the top and rows for personal information, documents and settings
Profile
The settings screen, with rows for notifications, language, appearance and security
Settings

A name

Nobody here knew what a money box was

Savings came to the US as “Money boxes”, a straight translation. Research showed the words weren’t the problem. The meaning was.

  1. BrazilCaixinhas

    Little boxes. People split their cash into envelopes to keep themselves honest.

  2. Mexico and ColombiaCajitas

    The same idea in Spanish, and the same habit behind it.

  3. United StatesPiggy Bank

    One place that is mine, and a memory of saving coins as a kid.

Koto created Nu’s 2026 brand system. I directed how it was applied in the US: illustration, imagery and photography, the Piggy Bank included.

Craft

The details I’m proudest of

  1. Marketing screen templates

    One template system behind cross-sell, Journeys, offers and welcome banners, so each team could ship a new message without a custom screen.

  2. Pull to refresh on Home

    The most repeated gesture in a banking app, and the moment the new Home feels alive.

  3. Loading Settings

    How the profile and account details load, worked out hand in hand with engineering over several rounds until it held up.

The welcome screen, with a country switch at the top right
The profile screen, with a country switch at the top and Log out at the bottom

One product, more than one country

Nu has one account per country, and some of our first US customers already had one in Brazil, Mexico or Colombia. To them it was one product with several logins, and that was confusing.

  1. 1A country switch on the welcome screen.
  2. 2The same switch at the top of the profile.
  3. 3A clear way to log out and into the other country.
Looking back

What went right, and what went wrong

Four things that worked and four that didn’t.

Went right

  1. I got every design lead in one room first

    In my first month I gathered every design lead the US depended on, across business units and markets, plus the design system team. We agreed on the patterns before anyone built.

  2. Testing settled the argument

    We tested the smart widgets with people in Brazil, then in the US. Both groups said the same thing, so the simpler Home had evidence behind it.

  3. The US went first

    Rather than adapt Brazil’s app, the US became the first product built entirely on the new patterns. That is why the other markets could adopt it.

  4. Everyone walked the product together

    With the Chief Design Officer I ran sessions where every executive, product manager, designer and engineer on the launch went through the live app, decided what to change and named an owner.

Went wrong

  1. The smart widgets confused people

    People thought we were going to take their widgets away and change their Home for them.

    What I changedEvery widget has three states and stays where people put it.

  2. “Money boxes” meant nothing here

    Savings came to the US as a straight translation. In research, people didn’t know what a money box was.

    What I changedIn the US it is the Piggy Bank.

  3. I ran the marketing team too

    I stood up the US marketing team and led it alongside the product until the load was too much.

    What I changedI brought in a dedicated lead from outside the launch.

  4. We simplified a bit too much

    The Home came out practical when it could also have been expressive.

    What I’d changeKeep the structure, and spend the extra weeks making each widget feel like yours.

For your team

What I’d bring to your team

At Nubank I used Claude and Cursor to prototype product ideas. On Panols I took my own idea all the way to the App Store.

See all my workNext: Healthy Together

  1. 1A brief that people and AI both work from
  2. 2Your design system, written as rules AI can follow
  3. 3Nothing counts as done until someone has watched it run