>> about me

I build things.

I've been building since I was a teenager competing in programming olympiads, and twenty-something years later the compulsion is the same. The problems are just more interesting now.

Origin

I've been coding since high school. Competed through provincial and national rounds of the Vietnamese Informatics Olympiad, picked up a third-place finish at the national level. The real legacy wasn't the medal though. It was a 1.44MB floppy disk packed with Pascal and Assembly: algorithm solutions, a Tetris clone, a Norton Commander clone, even a Windows 3.1 style UI running on top of DOS. That training shaped how I think about systems: constraints first, then the simplest structure that handles them.

From there: software engineering at Vega, then engineering leadership, then CTO. Twenty-plus years building consumer internet products in Vietnam.

The through-line

At Vega, I led Clip.vn to Alexa #17 in Vietnam - ~20M monthly pageviews on horizontally-scalable infrastructure I designed, including a custom Vietnamese full-text search engine for MySQL years before native multi-byte support existed.

At Tiki, I joined as the 5th engineer and built the engineering organization from 5 to 100+. Led Talaria - the ground-up rewrite that replaced a heavily-customized Magento monolith with an in-house platform sustaining 15-20k orders/day and 20k req/s at peak.

At Viec Co, I co-founded Vietnam's on-demand staffing platform and grew it to 100,000+ pre-verified workers and 15,000+ monthly placements across 200+ enterprise customers - Shopee, Tiki, Lazada, Thế Giới Di Động, Bách Hóa Xanh. Raised Pre-Series A led by Access Ventures.

Now

I'm shipping a portfolio of multi-platform consumer apps across iOS, Android, web, macOS, and Windows - Strait Escape, GeoMinesweeper, TaskPaw, and more in active development.

Alongside that, I'm actively developing Code Agent workflows for the software development process - custom agents for code review and tech-lead loops, skill-based context that separates auto-enforced rules from on-demand methodology, sandboxed automation, and explicit safety boundaries. The goal: raise shipped quality and velocity at lower engineering effort.

I'm not trying to build the next unicorn. I'm trying to build things that work well, stay honest about what they do, and last longer than the trend that inspired them.

I still write code every day.

How I think

• No architecture is perfect. There's only the one that fits the business goal best.

• Design for scale, but ship the product first.

• Push your own limits before chasing trends. Otherwise you never actually learn anything.