Product Builder · 14 years in QA
강홍재 / James
After 14 years as a QA engineer, I bring that lens to product decisions - connecting failure modes, user problems, and release risk. I started by finding defects; today I work on what to build, why, and how to validate it. Open to the people and teams I can build alongside.
- QA Engineer
- 14 years across mobile, web, and embedded - managing release risk end to end.
- Product Builder
- Builds tools for release decisions, prioritization, and experiments.
- Open to
- People and teams to build with · PO/PM roles · AI workflow integration.
Positioning
Up close with product risk for 14 years.
Risk sense
The release-meeting seat where "is this safe to ship?" comes up most often. What could go wrong belongs in the feature design, not the post-mortem.
User problem
Tracker tickets are only a slice of user pain. The frustrations users never bother to file are the ones that should move the product backlog.
Release judgement
GO / HOLD starts with data and ends with a person. The score, the reasoning, and the dissent should live on one screen so the decision holds up later.
Validation loop
Testing isn't a pre-release ritual; it's a loop that narrows a hypothesis. The reproducible-experiment instinct a QA engineer has carries into product discovery.
Selected Work
Things I built.
ReleaseGate
MVPAn AI evaluation SaaS that turns release-or-hold decisions into data - just before launch.
2026 · Founder · Product · QA EngineerDocuStory
MVPAn analysis engine that turns Korean real-estate registry PDFs into risk signals - parser, rules, analyzer, platform, infra, and samples split across separate repos.
2026 · Founder · Product · EngineerFrameboard
LiveRICE · ICE · MoSCoW · Value × Effort in one workspace - open-source prioritization.
2026 · Founder · Product · Engineer
For founders without a builder
When you have the idea but no one to build it.
MVP 0→1, the next step for a service already in market, QA and release setup - I partner up to build. Send a one-page brief by email and we'll figure out together whether it's worth building.
If we should build something - or think it through together.
Collaboration, hiring, product feedback - all welcome. Replies usually arrive within 1-2 business days.