Chee Xing Wong

Mid-Level Software Developer @ Phillip Nova

About

I'm a full-stack engineer working across ASP.NET Core, React, PostgreSQL, and SQL Server. I build and ship multi-tenant production systems end-to-end — frontend, APIs, database, and the Kubernetes deployment that carries them. Most of my recent work has centered on backend performance and small-team delivery. I lead projects of one to three engineers, make the architectural calls, and own the parts most people avoid: connection pooling, caching strategy, test coverage, deployment pipelines. I care more about systems that stay healthy at 2am than ones that look clean on a whiteboard. The biggest shift in my work this past year has been rebuilding my engineering loop around AI. Not just running Claude Code alongside my IDE, but treating it as a system to engineer: custom project rules, scoped skills, orchestrated subagents, workflows that let a small team ship like a bigger one. I'm still growing into software architect and tech lead roles, but my bet is clear: the engineers who treat AI as a craft, not a shortcut, are the ones who'll define the next decade.

Country

Malaysia

City

Klang

Industry

Computer Software

Skill

PgBouncer, Redis, Claude Skills, Kubernetes, Git, Generative AI, LangChain, ASP.NET Core, xUnit, Docker, Entity Framework, GitHub, Clean Coding, JavaScript, PostgreSQL, C#, React.js

Experience

Mid-Level Software Developer

Phillip Nova

2026-4 - Present · 6 mos

Kuala Lumpur

Software Developer

AutoCount

2022-7 - 2026-4 · 3 yrs 10 mos

Klang

1. Progressed from full-stack engineer to tech lead across three major AutoCount platforms. Started on the QR Ordering system (2022–2024), grew into the tech lead role mid-project, then led an academic assessment platform (2025–present) and a CRM platform (2026–present) from the start, running teams of 1–3 engineers. 2. QR Ordering system — multi-tenant F&B self-ordering platform serving 100+ active restaurants in production. Owned architecture, API design, PostgreSQL schema, React frontend, and Kubernetes deployment. 3. Eliminated database connection pool exhaustion on the F&B platform by introducing PgBouncer and Redis-backed caching across polling-heavy paths. Reduced peak PostgreSQL connection usage by roughly 70 percentage points, stabilizing the system under sustained production load. 4. Academic assessment platform — designed an AI-assisted fallback layer on top of the rule-based marking engine. When strict answer matching can't grade a semantically-equivalent response (e.g., "F&B" vs "F and B"), AI evaluates correctness against prompt-defined rubrics, reducing false negatives and lecturer review workload. 5. Built role-based access control and admin impersonation into the same platform, letting the education team operate on behalf of lecturers during support flows without credential sharing. 6. Contributed core backend modules across other AutoCount products: subscription billing for a membership & loyalty platform, and enhancements to a cloud accounting platform. 7. Built out integration and unit test coverage with xUnit, covering API flows end-to-end before production. 8. Owned containerization and Kubernetes deployment pipelines across these systems, including integrations with third-party services and payment gateways. 9. Drove code review, architectural decisions, and adoption of modern practices around testing, observability, and AI-augmented development workflows across the team.