reliability · incident response · SLO · safe change

Reliability for systems that have to keep working.

I embed with teams as an SRE to keep their systems running. Incident response, SLOs, and safe change are the day-to-day.

Read the work →See the OSS

Zenn

SRE fieldwork, written up in Japanese.

Concrete write-ups of the incident response, SLO and safe-change work I do on client systems.

Zenn ↗

When Data Marts Proliferate and the Vendor Sends a Cost Increase Notice2026-06-15↗Why It Is Hard to Decide on Refactoring Proposals2026-06-01↗When Chasing Incidents, Experiences Fade Away2026-05-26↗Why Your Team Is Not Acting Even When SLOs Are Configured2026-05-25↗When Teams Fear Releases, There Is a Hidden Bottleneck No One Calls a Problem2026-05-18↗Stalled Schema Migrations in a 24/7 SaaS: Breaking the Three-Month Impasse2026-05-11↗

Tools

Tools I wanted on the ground, built as open source.

A personal OSS effort, kept separate from client work. I build what would speed things up on the ground and open-source it to make SRE better for everyone.

pgincidentGo

Real-time PostgreSQL incident dashboard for the critical first 30 seconds.

GitHub ↗

colrefGo

Find the real column references before you touch a Rails/Django schema.

GitHub ↗

gomarklintGo

Fast Markdown link & structure linter to keep docs from rotting.

GitHub ↗

tinytapC · eBPF

Kernel-level traffic visibility, built on eBPF.

GitHub ↗

see all →

the full toolbox on GitHub

dev.to

Introducing the OSS, in English.

Write-ups on how these open-source tools work and why I built them.

dev.to ↗

Couldn't you just use pg_activity? — staking out one corner of Postgres incident triage2026-07-01↗The first 30 seconds of a Postgres incident: why they take 30 minutes2026-06-23↗How to safely remove a Rails column: finding every real reference before you delete2026-06-16↗Renaming `user` when grep can't tell which model you mean2026-06-09↗How to safely remove a Django model field: finding every real reference before you delete2026-06-02↗How I Cut My Go Markdown Linter's Benchmark by 81%2026-05-26↗

Consulting

Let's talk reliability

Incident response, SLOs, safe change, observability. If something around the reliability of a long-running system is weighing on you, tell me what's going on. I start by getting a clear read on where things actually stand.

Let's talk