detour: Wander Together

A dating and friendship app for people on the move. detour matches digital nomads by route, proximity, and compatibility — not just location.

Comment

Inspiration

I chose Quinn Gable's creator brief because it hit close to home. I've spent the last few years living nomadically — Sri Lanka, Korea, Japan, Singapore — and I'm right at the point where I want to go full-time digital nomad. So when I saw the brief, I wasn't just building an app for a creator. I was building an app for myself.

And honestly? The current options out there are frustrating. Dating apps force you into endless swiping, lock basic features like seeing who liked you behind paywalls, and push super boosts that make the whole experience feel like a gambling machine. On the other side, nomad friend apps exist but they feel like an afterthought. Everything out there forces you to pick — dating or friends — with no middle ground. detour is that middle ground.

What it does

Detour is a dating and friendship app for digital nomads and van-lifers. Instead of just matching people by location, it connects people by route — where you're headed, not just where you are. It features proximity-based matching, a compatibility score system, and community features that help nomads find dates, friends, and help along their journey. Access is invite-only to keep the community intentional and high-quality.

How I built it

I started by downloading a bunch of the most popular dating apps and activity apps on the market. I went through each one trying to identify what worked and what didn't — essentially extracting the good and leaving the bad. I used Mobbin for design inspiration and took Quinn Gable's brief as my foundation, then layered in my own twist based on years of firsthand nomad experience.

For development, I used Cursor with Claude Code. I went through multiple AI coding tools throughout this process — from Bolt and Lovable to Claude Code to Codex — learning which models are best suited for specific tasks along the way.

Challenges I faced

The TestFlight curveball. I'm an Apple ecosystem guy — Mac, iPhone, the whole thing — so I was set on publishing via TestFlight. But this was my first time ever publishing an app, and I didn't know that being based in Korea meant I needed a Korean business registration number. On top of that, I didn't account for Apple's review process and its timing. I had to make a tough call: gamble on everything getting approved in time, or pivot to Android. I pivoted to Android.

Setting up RevenueCat. I'd never integrated a subscription platform before, so there was a learning curve. It wasn't insurmountable, but it took some figuring out as a first-timer.

Innovating in a crowded space. Dating and activity apps are not a novel idea — millions of people have tried. The real challenge was finding fresh angles. The compatibility score feature is something I'm genuinely proud of, but constantly pushing myself to think of features that feel new and different was tough.

AI-generated bugs. Building with Cursor and Claude Code meant I could move fast, but it also introduced unexpected bugs that were hard to even find, let alone fix. Debugging often took multiple prompts and rounds of testing. It sharpened my skills, but it was a grind.

What I learned

I came out of this hackathon genuinely confident that I can build an app from zero and publish it on both the App Store and the Play Store. That's a big shift.

My debugging skills leveled up significantly. My prompting skills got sharper. And I've developed a real sense for which AI models and tools are best for specific tasks — moving from Bolt/Lovable to Claude Code to Codex, each step teaching me something new about how to use AI to build software effectively.

Built With

Share this project:

Updates