Story of T-Apex: Zero Ideas, 24 Hours, and One Crazy Pivot to the Top 10.
Source: Dev.to
We’ve been here before. Last to last year, we walked into HackCBS full of confidence, and we walked out with nothing—not even a mention. It stung. So, when HackCBS 8.0 rolled around this November, we honestly went in with zero expectations. Actually, it was worse than that—we went in with zero ideas. We thought, “Here we go again, probably going to be worse than the last one.”
But man, were we wrong…
The Hackathon Experience
It was a 24‑hour hackathon, and for the first two hours, we (Team‑Apex) didn’t write a single line of code. We just sat there, staring at a whiteboard, planning the flow. The panic started to set in, but then we hit a rhythm. The whole night became a blur of VS Code, endless debugging, and enough caffeine to kill a horse.
The Turning Point
Around 3 AM we were stuck on a logic problem, and suddenly 2‑3 mentors swarmed around our desk. They weren’t just nodding; they were arguing, brainstorming, and genuinely excited. That’s when it hit us: we weren’t just building a normal app—we were building something unique. That validation fueled us more than the coffee ever could.
The Problem We Tackled
We tackled a problem every developer knows: the gap between a designer’s sketch and a developer’s code. Designers draw beautiful interfaces, and developers spend days fighting with CSS to make them look right. We wanted to fix that.
Why It’s the Real Deal
- It sees context – It doesn’t just see boxes. It uses vision to read your text. If you sketch “Nike”, it fetches the real Nike logo. If you sketch a food‑delivery app, it fetches food images.
- No templates – It writes code from scratch using Tailwind CSS (or other frameworks).
- Natural‑language refinement – Don’t like the color? Just type “Make it dark mode” or “Make it more professional,” and the AI rewrites the code instantly.
Results & Takeaways
We didn’t win the top 3, but out of hundreds of teams, landing in the Top 10 nationally is a massive win for us. From walking in with no idea to walking out with a working GenAI product, the journey was insane.
If you are hesitant about joining a hackathon because you think you aren’t “ready” or don’t have an idea—go anyway. The pressure‑cooker environment forces you to learn faster than any tutorial ever will. We learned more about multimodal AI in those 24 hours than we did in the previous six months.
Hackathons are really important, guys. The learning, the networking, and yes, even the debugging at 4 AM—it’s all worth it.