AI-Crafted App ArkFlow Finds Users Prefer Collecting Card Faces over Asset Analysis

Ark, who said this was his first article on SSPAI after more than 11 years on the platform, initially designed ArkFlow as a "life-freedom asset system" to help users understand how much living choice their assets could provide. The app included a 60-stage "freedom portrait" that analyzed assets, liabilities, income, expenses and goals. But after launch, almost no one discussed the freedom score or shared which character stage they had reached. Instead, the most frequent user requests were for uploading real card images, adding more stacking styles, hiding the homepage and analysis pages to leave only the card wallet, and building a wall to display collected bank cards.

"One user even told me he didn't care about asset analysis at all, and just wanted to use it to collect card faces," Ark wrote. He said this was a strong signal that developers often overestimate how much users value the most complex features. The card wallet, by contrast, required no explanation and gave users an immediate emotional response. As a result, Ark adjusted the product to make the card pack the primary interface, while keeping asset management and the freedom portrait in the background. Cards are displayed in three stacking modes: a traditional overlay, a focused view, and an Apple Wallet-style stack that expands and returns with detailed animations.

Ark said he had more than 50 financial accounts to manage, including over 10 savings cards, 20 credit cards, overseas bank cards, and investment accounts. His original pain point was not calculating total assets but remembering repayment dates, annual fees and where money was held. The app does not record every transaction; it focuses on the current snapshot of balances, credit limits, bill dates and currency conversions.

Privacy is anchored in the design. ArkFlow does not connect to bank accounts, does not require registration or online-banking credentials, and stores data on the user's device, with optional iCloud sync and Face ID protection. "As the developer, I don't know how much money a user has," Ark wrote. "For a finance app, not knowing is an important capability."

This was Ark's first deep use of AI to complete an iOS product. He said AI significantly shortened development time, but did not make solo development easy. AI could implement clear requirements quickly but could not judge product direction, feature trade-offs or interaction aesthetics. He encountered issues such as code that broke animations after changes and conflicts among separately completed features. He eventually divided tasks into separate Git branches and worktrees, letting different models handle implementation, review and regression before he made the final decisions.

Ark also noted that some people found the app unnecessary, especially those with few accounts. But a segment of users spent considerable time searching for card designs and re-adding discontinued cards, discussing which year's design to use and how to arrange the cards. He concluded that ArkFlow does not need to cover every financial scenario. "A grand product concept is not necessarily the best first step for users to enter the product," he said.