AI Features
Jetson uses AI to turn customer conversations into actionable insights.
Conversation classification
Every conversation is analyzed to determine:
- Category - Bug, feature, or no action
- Priority - High, medium, or low severity
- Topic - What the conversation is about
High-confidence items are classified automatically. Lower-confidence items are flagged for review.
Bug report generation
When Jetson identifies a bug, it generates a structured report:
- Clear problem description
- Steps to reproduce (when available)
- Error messages mentioned
- Severity assessment
- Number of affected customers
Feature request extraction
For feature requests, Jetson captures:
- What the customer wants
- Why they want it (the underlying need)
- How many customers have requested it
- Priority based on request volume
Duplicate detection
Jetson recognizes when multiple customers report the same issue:
- Similar conversations are grouped together
- Customer count increases on existing items
- No duplicate work items created
Pattern labels
When issues are grouped into patterns, Jetson generates specific labels like "Login Flow Errors" instead of generic names.
Custom instructions
You can guide how Jetson generates work items:
- Go to Settings > Products
- Find Work Ticket Instructions
- Add custom guidance (up to 1000 characters)
Example instructions:
- "Always include steps to reproduce for bugs"
- "Format acceptance criteria as checkboxes"
- "Include the feature category in titles"
Related topics
- Custom Work Ticket Instructions - Detailed guide on customizing AI output
- How Classification Works - Learn about the classification process
- Patterns - Understanding how issues are grouped together