Bernard Koissy
IT Service Desk Supervisor at the National Institute of Standards and Technology, where I run support operations for roughly 7,000 users across two enterprise desks. Outside of that, I design, build, and operate production software end to end — frontend, API, database, deployment, and the monitoring that catches it when it breaks.
Five projects, all currently running. Each page covers what it does, how it is built, and what it took to operate.
A collection manager for sports and trading cards, running in production against a catalogue assembled from seventeen upstream data sources.
A family chore and allowance tracker, rewritten from a Home Assistant integration into a standalone web app. GPL-3.0, source public on GitHub.
A self-hosted platform of roughly 78 containers across six hosts, with its own certificate authority, single sign-on, full observability stack and offsite backups.
Custom Model Context Protocol servers, a suite of self-healing automation workflows, and an automated weekly security pipeline — the tooling layer that operates everything else.
A pickup-sports app for organising games, currently in public beta — three separately deployed surfaces sharing one managed backend.
I have spent twelve years in IT operations, most of it in federal environments where uptime is not negotiable and the people you support are not technical. My current role covers two enterprise service desks: ten analysts, roughly 3,000 tickets a month, a 95%+ service level and an 85% first-contact resolution rate. I am ITIL and Azure certified.
The building side started as a way to understand the systems I was supporting and turned into something larger. I run a homelab of about 78 containers across six hosts — private certificate authority, single sign-on, metrics, logs, alerting, offsite backups, zero-trust remote access — and it is where every production pattern I use gets tried first. The projects below are what came out of it.
I am most useful where operations and engineering meet: making systems observable, making failures recoverable, and making the on-call rotation quieter than it was last quarter.
Open to conversations about IT operations leadership, platform and reliability work, or anything below.
