Everything as code
Infrastructure, pipelines, and policy live in Git — versioned, reviewed, and reproducible. No console clicks, no snowflakes.
// DevOps & Platform Engineer · Seoul
I make shipping boring
I build the platforms, pipelines, and guardrails that let teams deploy dozens of times a day without paging anyone at 3am. Reliability, automation, and cost you can actually see.
How I work
What a platform engineer actually buys you
Not more tools — fewer decisions. I turn tribal knowledge into paved roads teams can walk without thinking.
Golden paths, not gates
Self-service templates give teams a safe default for deploys, databases, and dashboards — so security is the easy path, not a blocker.
Measured reliability
SLOs, error budgets, and DORA metrics on every service. Each case study below lists the numbers that moved, not just the stack.
A deploy's journey on a paved road
git push
Commit
Trunk-based, reviewed, and signed — the only manual step in the chain.
~2 min
Build & test
Hermetic builds with cached layers; flaky tests quarantined automatically.
~5 min
Scan & policy
SAST, image signing, and policy-as-code checks that fail loudly and early.
~8 min
Progressive rollout
Canary to 5%, watch the SLOs, then automatic promotion or rollback.
always
Observe
Dashboards, traces, and error budgets wired in before the first request.
Selected work
Platforms & pipelines I've shipped
A handful of case studies from real production systems. Click any card to read the role, stack, and measured impact.
FeaturedInternal Developer Platform on Kubernetes
2024A self-service platform that lets product teams ship to production without filing a ticket — golden paths, not gatekeeping.
KubernetesArgo CDCrossplaneBackstage
View case study 
FeaturedZero-Downtime CI/CD Pipeline
2023A progressive-delivery pipeline with automated canary analysis that ships dozens of times a day and rolls itself back.
GitHub ActionsArgo RolloutsPrometheusCosign
View case study 
Unified Observability Stack
2023A cost-aware metrics, logs, and traces platform that gave every team one place to answer "is it my service?"
OpenTelemetryPrometheusGrafanaLoki
View case study 
Multi-Region Infrastructure as Code
2022A Terraform module library and drift-detection workflow that made a two-region active-active topology reproducible.
TerraformTerragruntAWSCheckov
View case study Got a platform that's slowing you down?
Whether it's flaky deploys, a noisy on-call, or a cloud bill nobody understands — I'm happy to dig in. Let's talk.