Currency Exchange, built into nobu Go

I inherited the project after the lead designer transitioned to another initiative. From that point onward, I owned the design direction, evaluating multiple interaction models, balancing product goals with engineering constraints, and refining the experience through three design explorations before delivering the final production-ready design for handoff.

I reviewed 7 Indonesian banking apps, and one pattern shaped everything: nobody put the full picture together. Both sides of the transaction, a clear live rate, and your account balance, all visible before you even start typing. That was the gap.

Designing the interface wasn't simply about fitting six components onto one screen. The real challenge was deciding which information deserved immediate attention and which could wait, without sacrificing compliance, user confidence, or transaction speed. Every design decision became a trade-off between clarity, trust, and simplicity.

The strongest part of this project was the messy middle. Each exploration failed in a specific way, and each failure pushed the design closer to something that actually worked.

I believed separating Buy and Sell into dedicated tabs would make the user's intent immediately clear, allowing them to focus on one transaction at a time. In practice, however, the transaction became fragmented. Users could only review one side of the exchange at a time, while the exchange rate and supporting information remained separated from the moment they were needed. Instead of understanding the transaction at a glance, users still had to interpret how everything fit together before continuing. Making the user's intent clearer didn't automatically create confidence. Reducing visual complexity made the flow easier to follow, but users still had to interpret too much before feeling ready to continue. Focus and understanding, it turned out, weren't the same problem, and solving one hadn't solved the other.

Instead of changing the interaction model, I explored whether enriching each step with more contextual information would improve confidence. Adding more contextual information reduced some uncertainty, but the transaction still felt disconnected. Key details such as the exchange rate, destination account, and source of funds became easier to find, yet they remained separated from the decisions they were meant to support. Adding more context improved understanding, but confidence depended on more than visibility. Related information needed to be organized as part of the same decision, not presented as separate pieces across the screen. If context alone wasn't enough, the open question was what it would take to make those pieces feel like one decision instead of several.

I believed organizing related information into self contained transaction blocks would make the exchange easier to understand without changing the familiar tab based interaction. Grouping related information made each side of the exchange easier to understand. Exchange rate, account details, and transaction context now worked together within each transaction block. However, users still had to switch between Buy and Sell to review the complete exchange, preventing them from understanding the transaction as a whole before continuing. Grouping related information made the transaction easier to follow, but the experience itself remained divided. As long as Buy and Sell lived in separate tabs, users still couldn't evaluate the complete exchange in one place before making a decision. Tabs were never the feature. They were the thing standing in the way of it.

Across all three explorations, the same problem kept surfacing: users were left to work out the things that mattered most on their own. The final design had one job, to make sure none of these questions ever needed to be asked.

The final direction balanced several competing priorities. The business wanted a faster, more direct exchange flow with fewer steps. Compliance required key information to remain visible throughout the transaction, while technical constraints limited features such as rate locking. Rather than hiding information behind tabs or additional screens, the final design kept the complete transaction visible from start to finish. The result was a denser interface, but one that reduced uncertainty at the moment people committed to the exchange.

This wasn't the cleanest interface I explored, but it was the one that best balanced user confidence, business goals, and technical reality. Not every improvement needed to ship on day one. The goal wasn't to create the perfect exchange experience. It was to create the right one for this stage of the product, knowing the design could continue to evolve as the product matured.

This project changed the way I make design decisions. I used to believe that removing elements naturally made an interface easier to use. Working through three failed explorations taught me otherwise. Simplicity isn't about showing less. It's about giving people enough clarity to move forward with confidence.

Like many designers, I naturally gravitated toward cleaner layouts because they often felt more elegant. This project challenged that instinct. If removing information makes people less certain about their decision, the interface may look simpler, but the experience becomes harder to trust.

Instead of thinking about how to make the interface cleaner or more visually appealing, I started asking what people actually needed to feel confident and reassured before continuing. Once that became clear, removing unnecessary elements became much easier because every design decision had a purpose.

That shift in perspective also changed how I evaluate my own work. Looking back, there are still a few decisions I'd approach differently today.

The lesson: every constraint left something out. The real work was deciding which gaps were acceptable to leave for later, and which ones weren't.