Lessons Learned Building SafeSpend

When I set out to build SafeSpend, the goal was simple: replace the anxiety of mental math with a single, clear "Safe to Spend" number. But as any developer knows, building something that feels simple is usually incredibly complex under the hood.

Here are the biggest lessons I learned while building a financial app specifically designed for the AuDHD brain.

1. "Minimal" UI is Harder Than "Feature-Rich" UI

My instinct as an app developer is often to add features. Graphs! Pie charts! Trend lines!

But every single data point on a screen increases cognitive load. During user testing with neurodivergent individuals, I realized that seeing a pie chart of last month's spending didn't help—it just triggered a shame spiral.

The hardest design decision wasn't what to include; it was what to ruthlessly cut. I had to kill the pie charts. SafeSpend works because it gives you the one number you actually need to know before you tap your card.

2. Open Banking APIs are Powerful, but Messy

Connecting to UK banks via Open Banking is a modern miracle, but transaction data is incredibly messy.

A coffee shop might appear as POS/TESCO EXPRESS 4022 on a Tuesday, and TESCO-STORES-4022 on a Wednesday. Trying to build predictive algorithms to identify recurring bills required a lot of trial and error. I learned that relying entirely on exact string matching is a fool's errand, and building fuzzy-logic categorisation is essential for calculating an accurate "Safe to Spend" number.

3. Pending Transactions are the Enemy of Clarity

The "Payday Gap" anxiety is heavily fueled by the discrepancy between your "Current Balance" and your "Available Balance." Banks often hold pending transactions in limbo for days.

Handling pending transactions was one of the biggest technical hurdles. SafeSpend had to be engineered to aggressively deduct pending charges immediately, because for an ADHD brain, if the money is visible in the account, it feels like it's available to spend.

Conclusion

Building SafeSpend taught me that accessibility isn't just about screen readers and color contrast. Cognitive accessibility is about protecting the user's working memory.

We need more apps that do the heavy lifting silently, presenting only the absolute truth.