Resolving an SLA Breach or Exception
The five-step exception-resolution workflow from Operations Command.
Overview
Every exception surfaces in Operations Command with an SLA timer, an owner and a suggested action. Resolving one is a five-step loop that always ends with a customer communication and an audit entry.
Business Value
Predictable, teachable exception handling.
Every resolution generates a citable audit event.
Prevents SLA breaches escalating into customer-tier downgrades.
How It Works
From Operations Command, open the exception, review the AI recommendation on the Shipment Twin, take corrective action, draft the customer update and close the task.
Workflow
- 1
Open the exception
Click the amber or red row in Operations Command's Exception Feed.
- 2
Read the AI Context Rail
On the Twin, read the risk headline, drivers and next-best-action — every claim is cited.
- 3
Take the corrective action
Approve the recommended action (reroute, escalate, rebook) via the approval dialog.
- 4
Draft and send the customer update
Use the Customer Update Drafter — the message is pre-filled from Twin state.
- 5
Close the task
Mark the exception resolved. The audit trail records actor, action, citation and timestamp.
Key Components
Best Practices
- ▸Never close an exception without sending the customer update — the audit trail expects it.
- ▸Escalate to AOG Command the moment an aircraft becomes involved.
AI Insights
Recommendation, not automation
The Copilot recommends; the human approves. That boundary is what makes the audit trail defensible.
Frequently Asked Questions
What if I disagree with the AI recommendation?›
Override it and record the reason in the approval dialog — the alternative reasoning is captured in the audit trail.