🔍 Root Cause Analysis
AI Root Cause Analysis examines an incident's check history, error messages, and concurrent incidents to determine why something went wrong.
🔧 Como Usar
- Go to the Incidents page
- Find the incident you want to analyze
- Click the "AI Root Cause" button (purple sparkle icon)
- Wait a few seconds for the AI analysis
📊 What You Get
| Campo | Descricao |
|---|---|
| Root Cause | The most likely reason for the incident |
| Confidence | High, Medium, or Low |
| Category | network, server, dns, ssl, application, database, third_party, infrastructure, configuration, unknown |
| Evidence | Data points supporting the conclusion |
| Recommendations | Steps to prevent recurrence |
| Related Analysis | Whether concurrent incidents are connected |
🧠 What the AI Considers
- Check history 30 minutes before and during the incident
- Response times, status codes, and error messages
- Other incidents happening at the same time on the same team
- The monitor type and configuration
- Duration and pattern of the outage
💡 Dicas
- More check data = better analysis. Monitors with shorter intervals give the AI more data points.
- Run this after an incident is resolved for the most complete analysis.