Why we stopped tuning timeouts by feel
A short writeup on moving from guessed timeout values to ones derived from actual latency percentiles, and what changed in our incident rate afterward.
We build infrastructure that quietly stays out of your way.
Norrvik Labs is a two-person studio working on backend systems, deployment pipelines, and the unglamorous plumbing that keeps services running at 3am without anyone noticing.
What we work on
Service architecture
Designing backend systems that fail predictably and recover on their own — queues, retries, idempotency, the boring stuff that matters.
Deployment & CI
Reproducible build pipelines, container images, and release processes that don't require tribal knowledge to operate.
Observability
Logging, metrics and tracing setups sized for small teams — enough signal to debug an incident, not enough to drown in dashboards.
Hardening & ops
Server configuration, access control, and the small habits that keep a production box boring for years at a time.
Notes
A small checklist before handing off a server
The handful of things we verify — backups, access, monitoring, documentation — before considering any box "done" and moving on to the next project.
Notes on running services on modest hardware
Most small projects don't need a cluster. A writeup on getting comfortable, predictable performance out of a single small VPS.
About
Norrvik Labs is run by two engineers who have spent the last several years building and operating backend systems for small teams. We take on a handful of contracts at a time, prefer long email threads over meetings, and generally try to leave things simpler than we found them.