I usually spend my weekends working on my iOS apps, visiting hundreds of cafes in and around where I live, or meeting friends. Last weekend was different. I joined a hackathon in Chiang Mai, Thailand and competed against 27 other teams.
The Cursor event was three hours. Build something from scratch, ideally using sponsor tools like ElevenLabs.
I’m a 51 year old entrepreneur who learned iOS development last year using AI tools like Cursor. No computer science degree. No traditional coding background. No team. Just me.
By the end of the afternoon, I’d won. Beat all 27 teams. The project was Aphasio, a communication tool for those suffering from aphasia who lose their ability to speak, often after a stroke.
Inside the Competition
The hackathon ran on a weekend afternoon in Chiang Mai. Participants ranged from experienced developers to newcomers, most working in teams of two to four people. The constraint was the challenge. Three hours forces brutal clarity. You can’t build everything, so you build what matters most.
I’d been thinking about aphasia communication tools for months. Someone close to me had a stroke and developed aphasia. They understand everything but can’t form words depending on severity. What they needed was easy way to practice everyday and fast access to basic needs like “I’m hungry” or “I need help.”
I knew exactly what to build. The hackathon just gave me the deadline to ship it.
What I Built
Aphasio is a speech practice app for people with aphasia. The interface shows common phrases organized by category: greetings, daily needs, emotions. Users can listen to phrases spoken by an AI coach, practice speaking them, and track their progress. The app uses ElevenLabs voices for natural-sounding speech and provides encouragement as users work through their practice sessions.
I built the core functionality in three hours. Multiple phrase categories, AI voice coaching with ElevenLabs integration, practice tracking, and a clean interface designed for people recovering from stroke. Not fancy, but functional. Ready to use.
Why Solo Won Against Teams
Watching teams struggle with coordination while I kept shipping was instructive. Here’s what gave me the advantage:
Speed of decision making. Teams debated tech choices, divided tasks, argued about features. I made decisions in seconds and moved on. By the time most teams agreed on their approach, I had a working prototype.
No coordination overhead. Teams spent time explaining ideas to each other, syncing code, resolving merge conflicts. I wrote code, tested it, fixed it, moved forward. Every minute went into building.
Using the right tools. I built in Swift, a fast compiled language perfect for iOS apps. Combined with Cursor for AI-assisted coding, I wrote production-quality code at absurd speed. Most teams were wrestling with web frameworks or unfamiliar languages.
Clear problem definition. I had 17 years of business experience identifying problems and building solutions. I knew what features mattered because I’d watched my mother struggle with existing apps.
Deep tool expertise. I’ve used Cursor extensively for months. I know how to prompt it, when to trust it, when to override it. I wasn’t learning the tool during the competition. I was using it like a musician uses their instrument.
Major Takeaways
The hackathon crystallized something I’ve suspected for months. AI coding tools haven’t just lowered the barrier to building software. They’ve changed what matters.
Product sense beats coding skill. Half the teams had better engineers than me. But they didn’t have better problem definition. They built technically impressive projects that solved vague problems. I built something simple that solved a specific, painful problem. The judges responded to clarity.
Seventeen years of business experience mattered more than a CS degree. I’ve spent nearly two decades watching products succeed and fail. I know what users actually need versus what developers want to build. That instinct, that taste, that’s harder to acquire than coding skills.
Solo can beat teams when the tools are good enough. This wasn’t possible five years ago. Maybe not even two years ago. But AI-assisted development has compressed the advantage teams had. For certain projects, a single person with clarity can move faster than a coordinated group.
The barrier dropped but the bar raised. Anyone can code now. Cursor, Claude, other AI tools handle the syntax and boilerplate. But knowing what to build? Understanding users? Identifying real problems? That’s harder. The competition shifted from “can you code” to “do you know what’s worth building.”
Personal problems make the best products. I had an unfair advantage. I wasn’t building for a hypothetical user. I was building for my mother. I knew every pain point because I’d watched her experience them. Teams were designing for personas. I was solving for a person.
What This Changes
The landscape shifted. Startups need smaller teams. Solo developers can ship real products. Corporate dev teams will shrink. The question isn’t whether you can write code anymore. It’s whether you understand the problem deeply enough to build the right solution.
I’m using this win to establish credibility in the AI-assisted development space. Teaching others how to build with these tools. Showing that product sense and execution speed matter more than traditional credentials.
The hackathon proved what I suspected: identify a clear problem, know what the solution looks like, and AI tools handle the rest. The hard part was never the code. It was always knowing what to build.