A technology company operating in Qatar, hiring remotely, just posted a vacancy for an IT Support Lead. Read between the lines of that job description and you find a more honest document than they intended: a list of things a human still does that a machine already could.
The Ticket Board Eats Itself
The biggest chunk of the JD — the part they put first, because it's what the role actually is day-to-day — is "monitor and manage ticketing systems and service boards, ensuring tickets are assigned appropriately and resolved within SLA timelines." That's a triage and routing problem. An LLM with function-calling does this today.
The pipeline is this: your ticketing system's webhook fires on every new ticket. The ticket body — "printer on floor 3 not working," "can't access the VPN," "Outlook keeps crashing" — goes to a classifier model. It categorizes by issue type, infers priority from language and user role, checks against SLA rules, and routes. Not to a human queue. To the correct automation. Password reset? Hit the identity provider's API. Software install? Push via MDM. VPN config issue? Run the diagnostic script and email the user the fix.
For the subset that genuinely needs hands-on intervention — a dead switch, a hardware failure — the AI creates the work order, assigns it to the on-site tech, and sets the SLA timer. It then monitors the timer. If the human tech is about to breach SLA, the AI escalates to whatever vendor or contractor handles physical hardware.
The IT Support Lead who currently spends their morning staring at a service board, redistributing tickets, and prodding technicians? Gone. The board manages itself. The SLAs enforce themselves. The system doesn't get bored, it doesn't play favorites, and it doesn't need a mid-year performance review to remember to follow up on unresolved tickets.
Root Cause Analysis Without the Root Cause Meetings
The JD asks for "root cause analysis for recurring issues and preventive measures." This is the part of IT support that sounds intellectual and justifies the "Lead" title. In practice, it means: the same ticket keeps appearing every Tuesday, someone has to notice that, dig through logs, and find the misconfigured cron job.
This is pattern recognition over structured and semi-structured data. You feed the last six months of resolved tickets, system logs, and infrastructure alerts into a retrieval-augmented model. The prompt is simple: "Find clusters of recurring tickets with the same resolution path and identify the underlying systemic cause."
The model finds the recurring Tuesday spike. It correlates the ticket spike with the log entries from the backup job that runs at 2 AM. It drafts the preventive measure — "adjust backup window to avoid resource contention with the morning login wave" — and files it for human approval. The IT Support Lead's analytical function collapses to a 20-minute weekly review of AI-generated recommendations. You don't need a mid-senior level human for a 20-minute weekly review.
The Mentorship Mirage
The JD wants this person to "provide mentorship, training, and technical guidance to enhance team capabilities" and "train and mentor the support team." This is where the posting gets sentimental. Mentorship in IT support is not a Socratic dialogue. It's: "Here's how you clear the print spooler. Here's how you reset a user's MFA token. Here's the runbook for the finance app crashing."
That's documentation. That's a knowledge base. And a knowledge base that an AI both writes and queries is strictly better than one a human maintains. Every time the AI solves a novel ticket, it updates the internal knowledge base. When a junior human tech encounters the same issue, they don't call the Lead. They ask the AI, which gives them the exact steps, calibrated to their access level. The Lead no longer has to repeat the same runbooks to juniors. The team gets "trained" every single day, automatically, by the accumulation of resolved cases.
Where It Breaks: The Hands and the Politics
Here's the catch, and it's a real one. The JD includes "hands-on technical assistance and troubleshooting support for complex issues." A language model cannot physically walk to the third floor and see that someone has unplugged the core switch to charge their phone. It cannot smell the burnt capacitor in the server room. Physical IT support remains stubbornly biological. You still need a human with hands, at least until your office is full of networked cameras and IoT sensors that let an AI "see" the physical environment well enough to diagnose it. For most companies, that's not happening soon.
The second break point is organizational politics. IT support isn't just about fixing technology. It's about managing the expectations of the CFO who is furious that their presentation won't load, and who will not accept "the AI is working on it" as an answer. The IT Support Lead is often a shock absorber for human frustration. They translate technical reality into diplomatic reassurance. AI is getting better at tone, but people still want a person they can complain to. People want to escalate to a person. They want to feel that someone with authority is suffering alongside them.
And then there's vendor management. When the ISP goes down and the office is screaming, someone has to get on the phone with a human account manager at the telecom and apply pressure. AI voice agents can make the call, but they can't leverage the implicit threat of "I'm the IT Lead and I'm personally going to make this a problem for you." Not yet.
The Verdict
The IT Support Lead doesn't vanish entirely. It splits. The analytical, triage, routing, SLA-tracking, and knowledge-management functions — roughly 80% of that JD — fold into an automated system. What's left is a person who walks around fixing physical things, absorbs executive anger, and yells at vendors. That's not an IT Support Lead. That's an IT Support Specialist with better soft skills.
The company that posted this vacancy is hiring for a role that is already a fossil. They want a human to sit on top of a queue because they haven't realized the queue can drive itself. Job postings show which roles AI is already eating after the fact, and this one is a dinosaur bone sticking out of the dirt.
If you're applying for this job, you're applying to be the last human to touch a ticketing system before the company figures out it doesn't need you to touch it at all. Take the job. But automate your own workflow on day one, and don't tell anyone.
