AI Product Engineer · Swingvy
Jinyoung Park / Software Engineer
Planning, architecture, development, QA, and deployment — across the product lifecycle.
I’ve built and operated backends and cloud infrastructure for games and HR services. Today, I bring that experience to building products with AI agents.
01A deliberate shift
In late 2025, I began studying AI-assisted product development and expanding my responsibilities. I wanted to work across the entire product lifecycle. Through discussions with my company, I transitioned from Back-end Engineer to AI Product Engineer.
With a wider scope, defining what to build, structuring the work, and verifying results became more important. I now build products with AI agents, drawing on my experience developing and operating services.
FOUNDATION
Experience building and operating services
LATE 2025
Learning AI product development and expanding my scope
NOW
A new role, agreed with my company, across the product lifecycle
02How I work
I clarify requirements and constraints, then structure the work in detailed documents. I invest in making the scope and context clear for AI agents.
I use AI for code generation and testing, while considering the impact on existing systems and how to evaluate the quality of the result.
I prioritize delivery, operational effort, and cost. I’ve integrated structural improvements into feature work and evolved infrastructure in stages.
03Selected work
Evaluating new technology, improving architecture alongside feature delivery, and evolving the systems that run the product.
01 / AI Q&A · Swingvy · 2024
I built a Q&A system based on Singapore’s labour laws when LLM capabilities and available resources were limited. Answer quality, latency, and cost all mattered.
THE APPROACH
I refined retrieval and built a foundation for collecting user feedback on answer quality. I also created an evaluation process to compare models on response speed, quality, and cost.
02 / Backend · Swingvy
I incrementally extracted leave management from a monolith into an independent module, organizing the work around the feature delivery schedule.
THE APPROACH
During feature planning, I identified boundaries based on related functionality and potential side effects. Migration stories joined feature epics, so feature QA also covered the migrated behavior.
03 / Infrastructure · Swingvy
I built and operated AWS and GCP infrastructure, prioritizing early delivery and improving the environment over time with cost and operational effort in mind.
THE APPROACH
I moved from direct EC2 deployments to automation with OpsWorks and Chef. As OpsWorks support ended and management overhead grew, I migrated all services to ECS.
04Experience
2016.10 — Present
Built and operated HR backends and cloud infrastructure. Now working across the product lifecycle with AI agents.
2014.02 — 2016.10
Developed backend APIs for LINE Rangers, a mobile game with over 10 million downloads.
2013.06 — 2013.12
Started my engineering career as an intern.
05Tech Stack
From tools for building products with AI to the backend and infrastructure that keep them running.
LET’S BUILD TOGETHER
I’m interested in AI-assisted product development, backend and infrastructure challenges, and teams thinking about all three.
I welcome hiring conversations and technical discussions.
