Bayes
Bayes is Monte Carlo’s AI teammate. It observes your data and AI systems, fixes what breaks, and continuously improves your monitoring.
Bayes works across Monte Carlo as one teammate. You can ask it questions and give it tasks in chat, and it also does work on its own: it triages new alerts, troubleshoots them, builds monitoring coverage, tunes noisy monitors and reviews pull requests.
Bayes used to be documented as the Operations agent, with separate triage, troubleshooting, monitor tuning, cost & performance, PR and PII agents. These are now skills of one teammate. Some settings, permissions and billing line items still use the earlier agent names.
Where to find Bayes
- In the Monte Carlo UI. Select Bayes in the left navigation, or open the Bayes chat panel from any page.
- On your alerts. Bayes triages alerts as they arrive and can troubleshoot any alert on request. See Triage and Troubleshooting.
- In Slack and Microsoft Teams. Bayes is available through the Monte Carlo AI app. See Slack AI.
- In your IDE. See Use Bayes from your IDE.
What Bayes does
| Skill | What Bayes does |
|---|---|
| Chat | Answers questions about your data and takes actions you ask for |
| Triage | Scores new alerts by incident likelihood and impact |
| Troubleshooting | Investigates an alert and finds its likely root cause |
| Monitoring plan | Builds monitoring coverage for a domain |
| Monitor tuning | Tunes monitors to cut alert noise |
| Cost & performance | Finds storage and compute waste in your warehouse |
| PR review | Reviews pull requests for data-reliability risk |
| PII detection | Finds and monitors PII in your warehouse |
| Reinforcement Loop | Finds issues in your AI agents and proposes fixes |
Ask Bayes
Ask Bayes questions and give it tasks in plain language:
- "Are there any alerts assigned to me in the past day?"
- "How's the health status of table X?"
- "What are some trends and patterns in my alerts?"
- "When was my table Y last updated?"
- "What assets and reports would this data issue impact?"
- "Can I deprecate table X?"
- "Add a comment to X alert."
All operations exposed by the Monte Carlo MCP Server are available to Bayes in chat. See the MCP Server documentation for full details.
When a request needs a deeper workflow, Bayes uses the matching skill. You don't need to go to each skill separately:
| Request | Skill |
|---|---|
| "Troubleshoot this alert" | Troubleshooting |
| "Triage these incidents" | Triage |
| "Analyze cost for this warehouse" | Cost & performance |
| "Find PII in my warehouse" | PII detection |
Connect Bayes to your other systems
Connect your own MCP servers so Bayes can read context from systems like your ticketing tool or wiki when you chat with it and when it troubleshoots an alert. See Custom MCP servers.
Use Bayes from your IDE
Bayes works inside Monte Carlo's UI. If you prefer working from AI coding agents like Claude Code or Cursor, Monte Carlo offers a separate path:
- MCP Server -- connects your IDE agent directly to Monte Carlo.
- Agent Toolkit -- pre-built skills that teach your agent to monitor data, triage alerts, validate code changes, and more.
Ask support questions
Ask Bayes questions about your Monte Carlo setup, integrations, monitors and more. If you need additional help or a deeper investigation, Bayes can open a support ticket for you. See Getting support.
FAQ
What permissions does Bayes have when I chat with it? What data can it access?
When you chat with Bayes, it runs with the same permissions as you. It cannot perform any action or access any data that you cannot. Skills that run automatically, such as triage, are covered on their own pages.
Updated about 3 hours ago