Questions you’d ask me anyway.

On V-Assist, I owned a premium feature the company had already marketed: the drone scans a QR code from the air, tracks it in flight, and lands precisely on top of it — all driven from the Android-based DJI remote control.

Close to release, it was failing. At high image quality, QR scanning lagged behind the live feed, so the drone couldn’t follow the code in time to land. Most people would have dropped the image quality and shipped. Instead I debugged the pipeline line by line and found the real bottleneck: the video codec and rendering path, not the detection.

Working from the codec documentation, I rebalanced decoding and rendering so the feed kept its quality while scanning ran in real time. The feature shipped on schedule, the drone lands on the mark without crashing, and it’s live in production today. That’s my pattern: find the actual cause, don’t trade away the product.

The one closest to my heart is Theatre, a multimedia app from my Power GP BD years that crossed 500,000 downloads on the Play Store (it’s no longer live). Half a million people choosing something you built — that feeling never gets old.

The most recent one is Football Predictor 2026. FIFA’s new 48-team format is algorithmically nasty: the Round of 32 depends on third-place results across all 12 groups, with 450+ valid bracket combinations. I built an engine that resolves the whole knockout tree in real time, by the letter of the FIFA regulations. It turned something that confuses everyone into something anyone can play with.

Because I’m a bridge. I understand non-technical people — clients, founders, stakeholders — and translate what they actually want into technical decisions. Twelve-plus years of shipping end to end means I don’t just write code; I architect, build, release and maintain.

Two more things I bring everywhere: I automate redundant work instead of tolerating it, and I stay calm under critical deadline pressure — finding the smart solution rather than the heroic all-nighter. Ask my last three teams about release week.

Simple rule: I decide, AI executes. Architecture, tech stack and library choices are mine. Coding agents accelerate the implementation — but I track exactly what they’re using, steer them away from the wrong library toward the right one, and shape the solution so it stays future-proof: fewer bugs, less repair work later.

I keep myself current with cutting-edge tooling precisely so I can direct it. The agents work to my decisions, not the other way around. Every line that ships is a line I’ve reviewed and can defend.

Three doors are open. First, senior/staff/lead Android roles — I’m in Singapore and ready to relocate; I’m looking for companies that support that. Second, product development for clients: you bring the idea, I build it. Third, equity-CTO partnerships: I take over the entire technical side of an early-stage startup for a percentage plus ownership when it raises.

The thread connecting all three is ownership. I do my best work when I’m accountable for outcomes, not just tickets — building products I stand behind, with the freedom to do it right.