THE DIBASHI DAILY

DAYEMPIRE came with a challenge I genuinely hadn't faced before: build a real-time trivia platform from scratch, fast, and make it actually fun to play. Not just a quiz app. A live, competitive experience where timing, fairness, and engagement all have to work together at the same time.

So I built swl.live. The core of it is a real-time trivia engine I designed to handle session orchestration, live scoring, and leaderboard updates without falling apart under pressure. We hit 1,000+ active users in the first week alone, and sustained roughly 2,000 daily gameplay sessions. That's not a number I expected so quickly, and honestly, it's what pushed me to take the reliability side of things very seriously.

One thing that wasn't obvious upfront: question quality degrades fast at scale. Once you're generating thousands of questions, duplicates creep in and integrity breaks down. I built a validation pipeline with historical tracking to catch and eliminate duplicates across 10,000+ generated entries. It's not glamorous work, but it's the kind of thing that separates a polished product from a sloppy one.

The platform is still growing. I'm currently driving the mobile expansion, designing APIs and shared business logic that will let us ship on iOS and Android with offline tolerance and performance on par with the web experience. The engagement loop is something I keep iterating on too, using real-time usage data to understand where players drop off and what keeps them coming back.

earlySTEMer sits right at the center of my professional journey. I've been with them for years, and it's been a place where I got to do a bit of everything, building real systems, making key decisions, and even teaching.

As the sole developer on the team, I had the freedom and responsibility to shape our entire tech stack. Not every choice I made was perfect, but the impact was real. I built the Virtual Sessions Management System (VSMS) that finally got us off spreadsheets and helped cut down missed sessions in a big way. Then came the Lead Management System, which helped us stay on top of client conversations and grow the business. I later built a Student Portfolio platform so the kids could easily show off their work, and most recently, I launched a real-time chat system that lets students connect directly with live tutors when they need help.

But that's just the engineering side. I've also spent a lot of time tutoring/teaching dozens of students how to code, from the basics of Python to building full websites. Watching them go from “I don't get it” to building their own projects has been one of the most rewarding parts of this whole experience.

earlySTEMer gave me room to grow, fail, build, teach, and learn.

BREAKING: Developer Makes Tour Booking That Actually Works!

Torrista was a doozy. Still is, honestly. Working with the team at Sterling gave me a whole new perspective on tight deadlines and fast turnarounds. I was brought in to rebuild their rough proof-of-concept Django app into a proper production-ready system that could scale and actually work.

Luckily, I was already used to making technical decisions under pressure. I chose Node.js for the backend to make the codebase approachable for any developer joining the project. But the more controversial choice was going with Hono.js over Express. At the time, it raised eyebrows. I was betting on Hono becoming the kind of framework I knew it could be, and that bet absolutely paid off. Hono wasn't just faster and easier to work with, it could run anywhere. With Docker in the mix, deploying the app became ridiculously smooth no matter the environment.

Beyond the backend, I also built an admin application and a lightweight CRM to help the internal team manage day-to-day operations and generate reports without needing to touch code. The platform continues to evolve. We recently added a vendor module, and working on it is still just as exciting as it was on day one.

You're probably wondering why a full-stack developer has a mostly frontend project on their portfolio. You're not wrong to ask. The Blanche Aigle project earned its spot here because of one deceptively simple challenge: get all green Core Web Vitals scores and nail the SEO.

Sounds easy, right? But I wanted to do it using a proper framework, not just handcrafted vanilla code. I went with Nuxt, yes, Nuxt, not Next. If you haven't looked into it yet, please do. Might just change your life.

Nuxt gave me everything I needed to deliver. Its server-side rendering gave us excellent SEO support out of the box, and modules like @nuxt/image meant I could still build fancy sliders and dynamic layouts without sacrificing performance. The result was a site that looked sharp, loaded fast, and ranked well.

This project was genuinely fun to work on. It reminded me how much of a difference good tooling and thoughtful decisions can make when it comes to user experience.