Brazuka

Moving to Germany means rules that depend on each other. You need an address before you can register (the Anmeldung), and registration has to happen within 14 days. The bank wants the registration. Sometimes the landlord wants the bank. The answers exist, but they are spread across German portals, English-only guides and WhatsApp threads that ask the same questions every week.

Germany ranks 50th of 53 countries for expats, and 66% of expats find its authorities difficult to deal with (InterNations Expat Insider 2024). I audited the alternatives: Make it in Germany, Settle in Berlin, the Ankommen app and Portuguese blogs such as Euro Dicas. None was in Portuguese, mobile-first and step-by-step at the same time.

I combined desk research (community data, Destatis, InterNations, search trends and the groups where Brazilians in Germany talk) with a bilingual survey of 11 questions. It went out on 26–27 April 2026 in the WhatsApp and Facebook groups where people already find each other. The April study used the first 20 responses; the form stayed open and reached 22, and every number on this page uses all 22.

Who answered shapes what the answers mean: 21 of 22 respondents have lived in Germany for more than a year, and nobody was still planning or in the middle of a move. These are recollections of the move, not live experience.

Paperwork was the pain. Documents and bureaucracy were the most stressful part, ahead of housing and language, and nobody picked visa type or health insurance. For answers, people went to Google and to friends already in Germany before official sites.

Information wasn’t the gap. Almost half felt overwhelmed by the amount of it, only 6 of 22 found it easy to understand and follow, and half stayed neutral on that.

What people asked for was specific: explanations of the documents they need, step-by-step checklists, and a guide in their own language. A step-by-step guide in Portuguese would give 12 of 22 more confidence, and as many would trust an app made by a Brazilian who went through the move over other websites.

Money is the softest signal. Five said yes to a premium plan and ten said maybe, depending on the content. That is stated intent, not measured behavior, which is why the validation plan below puts real paid conversion first.

Two open questions asked for the biggest difficulty and the one thing people wish they had been told. Language, housing, bureaucracy and the emotional cost of the move came up most. They appear here as themes, never as quotes.

I turned the findings into a persona, an empathy map and a journey map. On the map, the emotion curve hits its lowest point in the first weeks after arrival, right where the circular trap sits. Five insights became five How-Might-We questions, and those questions set the scope.

Each design principle is anchored to a survey result. If a feature couldn’t be traced back to the research, it was cut.

The information architecture is deliberately shallow: four tabs (Início, Guias, Comunidade, Perfil), never more than two levels deep, so an anxious newcomer always knows where they are. Onboarding asks three questions (visa, timing, city) and hands over a personal plan before asking for an account. The signature screen is the bureaucracy map, which keeps each step locked until the step it depends on is done. The circular trap becomes a straight line.

Behind it sits a small design system (color roles, Inter with Newsreader Italic, a 4px spacing grid, three elevation levels), hi-fi screens in Figma and a clickable prototype in Portuguese covering onboarding, the checklist, the map, community, a visa comparison, a cost calculator, templates and the Premium paywall.

For the Professional Practice course I turned the research into a Value Proposition Canvas and a Business Model Canvas. The model is a freemium hypothesis: core guides free, Premium at €4.99 a month or €39 a year, affiliate partners for early growth, and B2B onboarding for employers kept as a later option.

I narrowed the MVP to the two phases where the pain is sharpest, pre-arrival and the first weeks, in Portuguese only and for four cities. Then I wrote a 90-day validation plan with explicit kill criteria. Paid conversion leads the metrics, because stated intent usually overstates real behavior, and four pivots are committed in advance if it stays below 5%. It is a plan. It has not been run.

The brief ended with a prototype. I kept going. Between 30 June and 3 July 2026 I built a web app with Next.js, Supabase and Netlify: sign-up with email or Google, a personalized onboarding, a checklist that saves progress, a community feed and Premium gating through a payment webhook. It is live. Payments are still in test mode.

From 6 to 14 July I rebuilt it natively in SwiftUI for iOS 26. Data stays on the device by default. It adds on-device Portuguese–German translation, a Face ID document wallet, a scanner that explains German letters in Portuguese, and an Apple Intelligence assistant that points people to the right guide instead of inventing advice.

That native build became the foundation of Brazuka: the research had shown the pain was real, so the next step was a product people could download.

Brazuka puts the pieces in one place. A feed filtered by city and topic. A directory of Portuguese-speaking professionals and Brazilian businesses. Classifieds, events, private chat, and the everyday tools a newcomer needs.

The whole interface speaks Brazilian Portuguese, and the visual language borrows Brazil’s colors without using the flag: deep navy, green and warm yellow on calm surfaces. The icon is a single ribbon that starts in Brazil’s green and yellow and ends as an arrow in Germany’s black, red and gold.

The arrival checklist, called ‘Chegando’, follows where you are in the move. A toggle switches between still in Brazil and already in Germany, tasks are grouped by theme, and your progress earns a badge on your profile. Before writing code I explored the product in a 32-screen product brief and an interactive HTML prototype.

After launch I redesigned the screens that carry the most trust and revenue: the paywall, event detail and professional profile. Dark mode, Dynamic Type and 44 pt touch targets are built in, and state restoration means a half-written listing survives iOS closing the app in the background.

The iOS app, written in SwiftUI, is the source of truth. The Android app is a deliberate 1:1 port in Kotlin and Jetpack Compose: each screen names the Swift struct it mirrors, spacing and radii are copied rather than reinvented, and a custom continuous-corner shape recreates Apple’s squircles. Where the platforms differ, the swap is an explicit decision. Anyone reading either codebase can trace a screen to its twin, which keeps both apps easy to review and hand over.

Both apps share one Firebase project: Authentication, Firestore, Storage, Cloud Messaging and Crashlytics. Seven Cloud Functions send push notifications, alert admins about reports, clean up past events every night and remove personal data when someone deletes their account. Firestore and Storage rules enforce ownership and field-level edit limits, and every rule change runs against the local emulator before it ships.

Sharing a backend surfaced real cross-platform bugs. Profiles created on Android once failed to decode on iOS because of one missing field. I fixed it in both clients and in production data. Xcode Cloud builds on every push and archives every tagged release, with configuration injected at build time.

Brazuka is free. Brazuka Pro, at €4.99 a month or €39.99 a year, unlocks the full step-by-step guides (Anmeldung, Steuer-ID, Aufenthaltstitel, bank account, Krankenversicherung), copy-ready German letter templates and a cost-of-living calculator. Contacting professionals, businesses and sellers is free for everyone: the monthly contact limit was dropped in October 2026 because it pushed newcomers away.

On iOS the subscription runs on StoreKit 2. On Android it is one Google Play Billing product with monthly and yearly base plans, so both platforms offer the same subscription.

Brazuka went live on the App Store on 25 July 2026. Android requests started in the first week. By week three, most of the engaged leads from my build-in-public posts on Threads were on Android and couldn’t install it. So platform reach became the only priority, and I built a native port rather than a web app. Google Play approved it for closed testing on 15 September 2026, and it went public on Google Play in October 2026.

I track a short weekly report of downloads, users, organic listings and content, and I tag small releases often: four per platform between 28 September and 2 October 2026.

Solo project: research, UX/UI, iOS (SwiftUI), Android (Kotlin/Compose) and Firebase backend, built with AI agents.

Twenty-two people is a start, not a representative sample. And it has a blind spot: 21 of 22 respondents have lived in Germany for more than a year, so they described the move from memory, and nobody answered from the middle of one. Willingness to pay was stated, not measured.

Next I would recruit a larger group that includes people still in Brazil or in the middle of moving, run usability tests on the Anmeldung flow with them, and measure real paid conversion instead of intent.