Helping high schoolers find extracurriculars, and learning what happens when real users show up.
Founder and engineer, 2025 – 2026
Opportunities for high schoolers are scattered everywhere: school sites, club pages, random posts. Some students have someone showing them where to look. Most don’t.
EC-AI puts those opportunities in one place, recommends ones that fit each student, tracks what they do, and helps turn it into something they can actually use for college applications.
When I started, I honestly didn’t know what I was doing. I was scared to even use Git, and there were a lot of times where I almost gave up because every part of it felt harder than I expected.
I just kept going one problem at a time, and eventually it reached over 8 million impressions on social media.
I picked tools that let me move fast: Clerk for auth, React Native, and Supabase on the backend. At the time, I wasn’t really thinking about rate limits, cost, or what would happen if usage grew a lot.
Once more real users started showing up, all of that caught up with me. We started hitting limits, costs went up, and people started running into crashes.
I went back and rebuilt a pretty big part of it. I figured out where the bottlenecks were, migrated parts of the stack, and changed the architecture so it could handle more users without constantly breaking.
Clerk and Supabase were fine choices. What I got wrong was picking them without knowing their limits or how I’d move off them once I hit them.
I still like moving fast. But now, before I make an early architecture decision, I usually ask myself two things: what happens if 10x more people use this, and where is this most likely to break?
Home