A full teardown to swap one part
A field repair that went deep: a full teardown to reach and swap a single failed part.
A zero-downtime server refresh
New ESXi host, RAID 10, fresh VMs, domain-controller cutover by promote/demote, backups test-restored before anything old was retired.
Four hours down, one bad cable
Racks built and rebuilt to a method, cables cut to length, tagged, and documented, so any fault traces end to end.
Getting signal into a metal building
Got network signal into a metal concrete-plant building where running cable wasn't an option, with a wireless bridge.
documents
Passwordless logins for a kindergarten class
A teacher logging a class of twenty-one five-year-olds in by hand, every session. I worked it end to end and wrote it up: policy first, then identity, devices, and the people, a proposal they could act on.
Working with the site's IT, not passing the puck
When a Xerox fault turns out to be the network, out of scope, I work with the site's IT instead of passing the puck.