Escalation risk

Detect whether [customer name] is showing active escalation signals across recent communications.

Step 1: Pull recent interactions

Gather the last 60 days of interactions across:

• Meeting transcripts and notes
• Support tickets and cases
• Emails and other communications

Step 2: Look for escalation signals

Repeated unresolved issues

• The same problem, bug, or complaint is mentioned across multiple touchpoints (meetings, tickets, emails)
• The issue has been open for 14+ days without resolution
• Language indicates urgency or business impact: “blocking”, “production down”, “can’t move forward”, “critical”, “this is stopping us”

Executive involvement from the customer

• C-suite or VP-level contacts (CEO, CFO, CTO, COO, VP) engaging directly in support threads or emails
• An executive being cc’d on a ticket or joining a call that would typically be handled by an end user or admin
• Language like “I’m looping in our CEO”, “our CTO is asking about this”, “escalating to leadership”

Escalation language

• Explicit requests to speak with someone senior: “we need to escalate this”, “can we get your VP on a call”
• Threats or ultimatums tied to unresolved issues: “if this isn’t fixed by…”, “we’ll have to reconsider…”
• Frustration compounding across multiple channels simultaneously

Step 3: Classify

• Escalation detected: One or more escalation signals found — repeated unresolved issue with urgency language, executive involvement, or explicit escalation request.
• -: No escalation signals detected.

Return exactly one of: Escalation detected, -

Include the specific signal (e.g., “CTO cc’d on 3 consecutive tickets about the same payroll sync issue, open for 18 days.”) in the reason field.