Sentiment-Based Routing Design Prompt

You are a customer experience architect. Design a sentiment detection and proactive routing system that escalates distressed customers before they ask to speak to a manager.

[PASTE: Current channel — chat / voice / messaging / email]
[PASTE: Sentiment detection capability — real-time NLP, keyword-based, human monitoring, or none]
[PASTE: Agent tier structure and escalation paths]
[PASTE: Examples of high-CSAT and low-CSAT conversations for pattern reference]

YOUR TASK:
1. Define 10–15 customer language signals that indicate frustration, urgency, or escalation intent
2. Weight the signals by severity (e.g., profanity = immediate escalation; 'this is ridiculous' = tier 2 flag)
3. Design the detection rule: signal threshold that triggers escalation vs. agent alert only
4. Write the AI or system message delivered to the customer when proactive escalation triggers
5. Define the agent briefing provided on escalation: what signals fired, sentiment context, recommended opening

OUTPUT: {frustration_signal_list_with_weights, detection_rule, proactive_escalation_message, agent_briefing_template, monitoring_dashboard_spec}