From ad hoc to a plan: organizing a fleet-wide change by site

Merging a device list with a user export to work a rollout site by site, and using vendor support instead of fighting an MFA issue alone.

The endpoint removal project needed a plan, so I built one that matched how the company is actually laid out: by office.

I merged the device inventory with the user export and broke the fleet out by site, three locations plus remote staff. One site had about fifty-nine machines, remote had twenty-nine, another had twenty-three. Sorting the list that way changed everything. Instead of a single number of two hundred machines, I had three or four manageable batches I could work on specific days, with named users I could reach directly.

I started with the largest site and worked the rest in order. The progress felt real because each completed site dropped a whole batch off the list at once.

The other thing that week was a sign-in and multi-factor-authentication problem that kept prompting one user in a business application. Rather than guessing, I worked it with the software vendor's support, changed the MFA method, and kept the user informed the whole way. It turned out to be more about how the identity provider and the app talked to each other than about the user at all. I closed thirty tickets that week, including a batch of warehouse receipt items in one morning, and I proposed a structured learning plan for the systems I want to master next.

The transferable lesson is that a long list is not a plan. A list sorted by location, with named owners and scheduled days, is a plan. Grouping by site also made the communication easy, because I could talk to one office at a time instead of spraying messages at everyone.

Slow is smooth, smooth is fast. Batching the work by where people sit made a two-hundred-machine project feel like a handful of small ones.

← All writing · Portfolio