What this Zendesk agent handles
High-priority tickets get lost when escalation depends on manual spotting or queue reviews. This agent surfaces risky tickets fast, explains why they need attention, and lets your team approve the next step before anything changes.
The agent handles three main escalation scenarios:
- Urgency keyword detection. The agent scans ticket content for urgency keywords like "urgent", "critical", "down", or "broken". It proposes priority increase and routing to senior agents.
- VIP customer handling. For tickets from enterprise or VIP customers, the agent checks customer tier and proposes immediate routing to dedicated support teams.
- SLA breach risk. When a ticket is approaching SLA breach threshold, the agent flags it for escalation regardless of content.
How the agent works
The table below shows the trigger, inputs, outputs, and approval step so you can judge fit quickly.
| Agent stage | What happens | Output | Approval |
|---|---|---|---|
| Trigger | A ticket is created or updated with potential urgency signals | Escalation review starts | No approval for read phase |
| Inputs | Ticket content, customer tier, SLA time remaining, priority field | Urgency score and proposed route | No approval for data gathering |
| Proposed action | Increase priority, reassign to senior team, add urgent tag, notify manager | Prepared ticket change | Approval before any write |
| Operator review | Reviewer checks urgency signals and proposed routing | Approve, reject, or edit | Explicit approval required |
How to set up this Zendesk agent
Define your escalation criteria in plain language, including urgency keywords, customer tier rules, and SLA thresholds.
Connect Zendesk with access to the ticket views, customer data, and priority fields needed for escalation.
Choose who reviews each escalation type, from routine SLA risk to high-stakes VIP tickets.
Set notifications so managers see proposed escalations quickly during high-priority incidents.
Permissions and approval checks
- Read access should cover ticket content, customer tier data, SLA tracking, and priority fields.
- Write scope should be limited to priority updates, assignee changes, tags, and approved notification fields.
- Escalation proposals should include the urgency signals detected so operators can validate the reasoning.
- Approval views should show both the ticket content and the proposed escalation path before the change is applied.
Frequently asked questions
Does escalation happen automatically or does it require approval?
By default, escalation proposals surface for review before the ticket is reassigned or flagged. You can keep clear cases fast while requiring review for high-risk or customer-sensitive tickets.
How does the agent detect urgency?
The agent scans ticket content for urgency keywords, checks customer tier, monitors sentiment, and looks for SLA risk. It combines these signals to propose escalation when appropriate.
What happens if the agent escalates incorrectly?
With approval gates enabled, every escalation is surfaced before it writes. You see the ticket, the urgency signals, and the proposed routing. Approve or reject. Nothing escalates without confirmation.
Can the agent escalate based on SLA time remaining?
Yes. The agent can propose escalation when a ticket is approaching SLA breach, regardless of content. You define the threshold and the agent surfaces tickets at risk.