Appearance
Sentiment & Tone Analyzer
The Sentiment and Tone Analyzer detects the emotional tone and urgency of ticket communications, helping identify frustrated customers or escalation-worthy situations.
What It Does
The processor analyzes ticket content (including recent messages and notes) to produce:
- Sentiment -- Positive, neutral, or negative.
- Frustration level -- A score indicating how frustrated the customer appears.
- Urgency -- How time-sensitive the issue seems based on language cues.
- Tone indicators -- Specific emotions detected (e.g., angry, confused, grateful, anxious).
- Escalation recommendation -- Whether the tone suggests the ticket should be escalated.
This processor runs on both creation and updates, tracking sentiment changes over time. If a customer's tone shifts from neutral to frustrated across messages, the processor flags this change.
Example Output
json
{
"sentiment": "negative",
"frustration": 0.78,
"urgency": 0.65,
"tone_indicators": ["frustrated", "anxious"],
"escalation_recommended": true,
"reasoning": "Customer has sent three follow-up messages with increasingly urgent language about a production system outage."
}When to Use It
This processor is enabled by default. It is valuable for identifying tickets that need immediate attention, monitoring customer satisfaction trends, and triggering escalation workflows when frustration is detected.
Details
- Phase: Continuous
- Runs on: New and updated signals
- Default State: Enabled