Service desk
agent
Tickets triaged and categorised properly, level 1 requests resolved to your runbooks, and the queue accurate enough that your engineers open a ticket already understood.
Your engineers are the most expensive triage layer available
In most small MSPs there is no real triage. A ticket arrives, an engineer picks it up, and half the time it is a password reset or an access request that took two minutes and interrupted an hour of focused work.
The cost is not the two minutes. It is the context switch, and the fact that the engineer who could be doing project work is instead running the queue.
A dedicated service desk agent absorbs the volume: triage, categorisation, information gathering and everything covered by a runbook. Your engineers get a queue of genuine technical work.
What they handle
- Ticket triage — Every ticket categorised, prioritised and routed against your matrix, with the obvious diagnostic information gathered before an engineer opens it.
- Level 1 resolution — Password resets, MFA re-enrolment, access requests, licence assignment, mailbox permissions and standard provisioning — anything with a documented runbook.
- Information gathering — Screenshots, error text, affected users, time of onset and reproduction steps collected up front, so an engineer is not asking basic questions on day two.
- Queue hygiene — Statuses accurate, stale tickets chased, and "waiting on client" actually meaning that — so your SLA reporting is real.
- Time entry chasing — Engineers chased for time entries, so billing reflects the work rather than what anyone remembered.
- Alert triage — Monitoring alerts reviewed against your rules, noise suppressed, genuine issues escalated with context attached.
- Onboarding and offboarding — User provisioning and deprovisioning run to a checklist — the process where security gaps most often appear.
- Client updates — Clients kept informed on open tickets, which is what drives satisfaction scores more than resolution speed.
Where to draw the access line
The technical work travels further in IT than most industries. The boundary is privileged access and change authority, not capability.
- Standing privileged accessGlobal admin and domain admin should be scoped, time-bound and logged — never standing access for a service desk role.
- Production changesAnything outside a documented runbook goes to an engineer under your change process.
- Security incident decisionsContainment and response decisions require your senior engineers.
- Client contractual mattersScope, billing and contract conversations stay with your account managers.