🤖 AI Configuration
Location: Admin → AI Configuration (under System)
Full setup guide is in the AI Features Overview. This page serves as a quick reference for all settings.
⚙️ Provider Configuration
| Настройка | Описание | По умолчанию |
|---|---|---|
| Enable AI Features | Master on/off switch | Off |
| AI Provider | OpenAI, Anthropic, Gemini, or Ollama | OpenAI |
| API Key | Provider API key (not needed for Ollama) | — |
| Base URL | Override provider URL | Provider default |
| Model | Specific model name | Provider default |
| Temperature | Response creativity (0-2) | 0.3 |
| Max Tokens | Maximum response length | 2048 |
🎛️ Feature Toggles
Each AI feature can be independently enabled/disabled:
| Feature | Default | Description |
|---|---|---|
| Root Cause Analysis | ✅ | AI analyzes incident causes |
| Postmortem Generator | ✅ | Auto-generate postmortems |
| Incident Correlation | ✅ | Find patterns across incidents |
| Status Page Drafts | ✅ | AI-drafted status updates |
| Status Translations | ✅ | Translate status updates |
| Natural Language Monitors | ✅ | Create monitors from English |
| Smart Thresholds | ✅ | AI-suggested alert thresholds |
| Anomaly Detection | ✅ | Detect unusual patterns |
| Predictive Alerts | ✅ | Predict potential outages |
| Weekly AI Reports | ✅ | Automated weekly summaries |
| Log Analysis | ✅ | AI-powered log analysis |
| AI Assistant | ✅ | In-app Q&A assistant |
🔘 Test Connection
Click "Test Connection" to verify your AI provider settings. This sends a simple test prompt and confirms the response.