๐ AI Features Overview & Setup
Monitron SaaS includes 12 AI-powered features that make incident management smarter, reduce alert fatigue, and save you hours of manual work.
๐ค Supported AI Providersโ
Monitron is provider-agnostic โ choose the AI provider that works best for you:
| Provider | Models | Self-Hosted? | Cost |
|---|---|---|---|
| OpenAI | GPT-4o, GPT-4 Turbo, GPT-3.5 | โ Cloud | Pay per token |
| Anthropic | Claude Sonnet 4, Claude Haiku | โ Cloud | Pay per token |
| Google Gemini | Gemini 2.0 Flash, Gemini 1.5 Pro | โ Cloud | Free tier available |
| Ollama | Llama 3.1, Mistral, CodeLlama, etc. | โ Self-hosted | Free! |
Budget-friendly option: Use Ollama with Llama 3.1 for free, self-hosted AI. Performance is good for most features, though cloud providers give better results for complex analysis.
โ๏ธ Setupโ
Step 1: Configure AI Providerโ
Go to Admin โ AI Configuration (under System in the sidebar).
For OpenAI:โ
| Einstellung | Wert |
|---|---|
| Enable AI Features | โ On |
| AI Provider | OpenAI |
| API Key | sk-... (from platform.openai.com) |
| Model | gpt-4o (recommended) or leave blank for default |
For Anthropic (Claude):โ
| Einstellung | Wert |
|---|---|
| AI Provider | Anthropic |
| API Key | sk-ant-... (from console.anthropic.com) |
| Model | claude-sonnet-4-20250514 or leave blank |
For Google Gemini:โ
| Einstellung | Wert |
|---|---|
| AI Provider | Google Gemini |
| API Key | From aistudio.google.com |
| Model | gemini-2.0-flash or leave blank |
For Ollama (Self-Hosted):โ
| Einstellung | Wert |
|---|---|
| AI Provider | Ollama |
| Base URL | http://localhost:11434 (or your Ollama server) |
| Model | llama3.1 (or any model you've pulled) |
| API Key | Not needed |
Step 2: Test Connectionโ
Click "Test Connection" to verify your AI provider is working.
Step 3: Enable Featuresโ
Toggle individual features on/off as needed. Each feature can be independently enabled.
๐ Feature Overviewโ
| # | Feature | Where It Appears | Description |
|---|---|---|---|
| 1 | ๐ Root Cause Analysis | Incident actions | AI analyzes why an incident happened |
| 2 | ๐ Postmortem Generator | Incident actions | Auto-write blameless postmortems |
| 3 | ๐ Incident Correlation | AI Assistant page | Find patterns across incidents |
| 4 | ๐ข Status Page Drafting | Incident actions | Draft public status updates |
| 5 | ๐ Translations | Status page drafts | Translate updates to other languages |
| 6 | ๐ฌ Natural Language Monitors | AI Assistant page | Create monitors from plain English |
| 7 | ๐ Smart Thresholds | Monitor actions | AI-suggested alert thresholds |
| 8 | ๐ฎ Anomaly Detection | Monitor actions | Detect unusual patterns |
| 9 | โ ๏ธ Predictive Alerts | Automated (scheduled) | Predict outages before they happen |
| 10 | ๐ Weekly Reports | AI Assistant page + email | AI performance summaries |
| 11 | ๐ Log Analysis | AI Assistant page | AI-powered log analysis |
| 12 | ๐ค AI Assistant | AI Assistant page | Ask questions about monitoring |
โ๏ธ Advanced Settingsโ
| Setting | Default | Description |
|---|---|---|
| Temperature | 0.3 | Lower = more focused, consistent. Higher = more creative. |
| Max Tokens | 2048 | Maximum response length. Increase for longer postmortems/reports. |
| Base URL | Provider default | Override for Azure OpenAI, proxies, or custom endpoints. |
๐ฐ Cost Considerationsโ
AI features make API calls that cost money (except Ollama). Estimated costs per action:
| Feature | Tokens (approx) | GPT-4o Cost | Claude Cost |
|---|---|---|---|
| Root Cause Analysis | ~2,000 | ~$0.02 | ~$0.02 |
| Postmortem | ~4,000 | ~$0.04 | ~$0.04 |
| Incident Correlation | ~2,000 | ~$0.02 | ~$0.02 |
| Status Draft | ~500 | ~$0.005 | ~$0.005 |
| Weekly Report | ~4,000 | ~$0.04 | ~$0.04 |
| Predictive Check (per monitor) | ~1,000 | ~$0.01 | ~$0.01 |
Typical monthly cost: For a moderate setup (50 monitors, 10 incidents/month, weekly reports), expect roughly $2-5/month with OpenAI GPT-4o.