What is AI Triage
AI triage takes raw, unstructured feedback — bug reports, feature requests, user complaints — and uses an AI model to convert it into structured tickets. Instead of manually parsing feedback and creating tickets, you can feed the raw input to the AI triage system and review the results.
How It Works
- Submit feedback — paste raw feedback text into the triage interface
- AI processing — the configured triage persona analyzes the feedback, extracts key information, and drafts one or more tickets
- Review — review the AI-generated tickets before they're created. You can edit titles, descriptions, priorities, and tags
- Approve — accept the tickets to add them to your project
Configuring Triage
To use AI triage, you need:
- An AI provider and model configured on your server
- A triage persona assigned to your project (see Providers & Models)
Best Practices
- Always review AI-generated tickets before approving — the AI may misinterpret ambiguous feedback
- Provide detailed feedback text for better results — more context helps the AI create more accurate tickets
- Use a dedicated triage persona with a system prompt tuned for your project's domain