The People Who Keep the Lights On
When the last crane leaves and the ribbon is cut, the workforce story does not end. It shifts to a different and quieter cast: the operations and technical staff who will run the facility every hour of every day for its entire life. This lesson is about who they are and why, person for person, they may matter even more than the army that built the place.
Start with the shape of the work. A data center is a live, continuously operating industrial facility that cannot simply be switched off for maintenance. It runs on a twenty-four-hour, seven-day basis, which means the staff must cover every shift, every weekend, and every holiday. Even a modest facility therefore needs enough people to fill a round-the-clock rotation with margin for vacations, sickness, and training. This is a small team compared to the construction peak, often a few dozen to a couple hundred depending on size and design, but it is permanent, and every one of those seats has to be filled by someone competent.
Who are these people? At the core are electrical and mechanical technicians who understand the facility's power and cooling systems intimately. They monitor and maintain the switchgear, uninterruptible power supplies, generators, chillers, pumps, and increasingly the liquid cooling systems that serve dense computing. They are supported by controls and building-management specialists who run the software layer that watches thousands of sensors and orchestrates the physical plant, and by critical-facilities engineers who understand how all these systems interact under stress. In many facilities there is also a separate IT and network operations staff, but the workforce that keeps the physical infrastructure alive, the power and the cooling, is the one that most directly determines uptime.
The consequence carried by each of these workers is what makes them so important. In construction, a mistake usually costs time or money and can be corrected before the building is live. In operations, a mistake can take live capacity offline in seconds, interrupting services that customers are paying for and depending on. A technician who mis-sequences a maintenance procedure on a power system, or who fails to catch a cooling anomaly before it forces a shutdown, can cause an outage that ripples out to thousands of customers. Conversely, a skilled operator prevents dozens of incidents that no one ever hears about, catching the failing component, the drifting temperature, the subtle alarm that precedes a real problem. The value of a great operations team is largely invisible precisely because it manifests as things that did not happen.
This is why the industry treats operational competence with such seriousness. Facilities run detailed standard operating procedures and methods of procedure for every significant task, so that maintenance on live systems follows a scripted, reviewed sequence rather than improvisation. New technicians are trained not just on the equipment but on these procedures and on the discipline of working safely around energized systems and stored energy. Building that competence takes time and mentorship; a technician may be technically credentialed on arrival but still need months to a year of site-specific experience before they can be trusted to run a shift on their own. The knowledge is partly general and partly specific to the particular facility's design and quirks.
Because the team is small and the expertise is deep, operations staffing is fragile in a way construction is not. Losing a handful of experienced technicians to a competitor can meaningfully degrade a site's reliability, because there is no bench of hundreds to absorb the loss. Each departure takes site-specific knowledge out the door and forces the remaining staff to cover more, which raises fatigue and error risk. In a market where several operators compete for the same small pool of skilled technicians, retention becomes as important as recruitment, and the employers who invest in training, career progression, and reasonable working conditions hold their teams while others churn.
The lesson to carry forward is that a data center is only as reliable as the people running it, and those people are a scarce, slowly-built resource just like the trades who constructed the building. Financing, design, and equipment get a facility to the point of operation, but it is the operations workforce that determines whether it actually delivers the uptime its customers are paying for, year after year. Anyone evaluating a facility's long-term prospects should look as hard at its ability to staff and retain skilled operators as at its power and connectivity, because in the end it is people who keep the lights on.