Category, severity and title โ automatically
Every incoming incident is classified in seconds: the AI assigns a category, estimates severity and writes a meaningful title. If it is unsure about the category, it deliberately leaves it empty instead of guessing. This removes manual triage and makes incidents instantly searchable.
- โCategory, severity and title automatically
- โWorks for app incidents and inbound incidents
- โCategory and severity can be overridden at any time
Photo analysis with vision AI
A picture often says more than a brief description. If an incident is described only briefly or not at all, the vision AI analyzes the attached photo instead. That way incidents with little text are classified too โ where the description is detailed, the text is enough.
Incident classified โ workflow starts
Categorization is not the end but the trigger: based on category and severity, your workflow rules fire automatically โ for example an alert on critical incidents. Where your data is processed is explained on our secure AI page.
FAQ
Frequently asked questions about AI categorization (FAQ)
Which AI is behind this and where is my data processed?
Classification runs on Azure OpenAI in a European processing environment. Processing is GDPR-compliant. Learn more on our secure AI page.
Can the AI also analyze photos?
Yes. If the description is brief or missing, the AI analyzes the attached photo to classify category and severity. Where the description is detailed, the text is enough.
What happens after automatic categorization?
Once an incident is classified, it can fire your workflow rules โ for example an escalation on high severity or a notification to the responsible person. You define which rule applies.
Can I override the AI suggestions?
Yes. Category and severity can be adjusted manually at any time. The classification then counts as human-confirmed and is not overwritten by the AI again.