Ongoing Remit
Technology & IT Operations
The company's networking, access and device management. The part of the remit nobody asks about until it stops working.
- Offices and warehouses
- Company-wide MFA
- Access control from HR data
- Apple device management
This is the part of my remit that has no launch, no dashboard and no applause. It is also the part that stops the company when it breaks, which is why it ended up with someone who treats it as a system rather than a queue of tickets.
It arrived the way these things usually do: the creative operation depended on infrastructure nobody owned end to end. Rather than route around it, I took it on.
How a new starter gets working
- HR record created
- Access derived from role
- Device enrolled
- Apps appear behind one login
- Leaver: access removed automatically
Networking across sites
- Standardised networking across offices and warehouses, so a site is set up the same way every time instead of inheriting whatever the last person configured.
- A warehouse is not an office. Coverage has to work around racking, cold areas and equipment that does not move, and the people using it cannot stop working while it is fixed.
- Standardisation is the point: when every site is built the same way, a fault is diagnosable remotely instead of requiring a flight.
Identity, access and MFA
- Company-wide MFA and access control, so an account is not one reused password away from everything.
- Access is derived from HR data rather than maintained by hand. Manual access lists always rot. Someone changes team, nobody updates the list, and the gap only surfaces during an audit or after an incident.
- Joiners get what their role implies. Leavers lose it automatically. Guest access expires on its own rather than depending on someone remembering.
- Elevated rights are grantable for one system, for a limited time, and withdrawable at any point.
Devices
- Apple device management across the company, so a laptop is a known, enrolled, recoverable asset rather than a personal machine with company data on it.
- Enrolment, configuration and policy are applied centrally, which is what makes offboarding a procedure instead of a negotiation.
Reliability as an operating habit
- Nothing ships from someone's laptop. Changes go through review and automated checks before they deploy. Internal applications get the same discipline as anything customer-facing.
- Systems fall back gracefully when something upstream is down and say the data may be stale, rather than showing an error and losing the user's trust.
- There is one register of what exists, who owns it, and where the documentation is, because "who owns this?" always gets asked at the worst possible moment.
How I approach the work
- Every one of these is the same move: find the thing that quietly fails, decide the rule that prevents it, make the rule automatic, then own the result.
- None of it is glamorous and all of it compounds. A company that standardised its networking, derives access from HR and enrols its devices has removed an entire class of recurring problem.
The point
IT operations is where the systems thinking gets tested without a launch to hide behind. It either works on a Monday morning in a warehouse or it does not.