Changelog
ChangelogStatus PageAPI ExplorerAPI ReferenceCLI ReferenceBlog
Back to All
Added

Analytics API (Public preview): report on the health of the data behind your agents, from your AI tool

about 7 hours ago

Trusting your agents starts with knowing the data that powers them is in good shape: which tables are healthy, what's alerting, where coverage is missing, and what it all costs. The Analytics API brings that picture into the AI tool you already use.

Ask Claude, ChatGPT, Cursor or any AI tool connected to the Monte Carlo MCP server, and get the dashboard you need on your Monte Carlo data, sliced by domain, data product, warehouse, tag or anything else in the catalog.

Your dashboards in Monte Carlo cover the everyday view. When you need a different breakdown, alerts and credits side by side, or a view for someone else, just ask the AI tool you already use.

  • Ask, or start from an example: Ask in your own words, or start from one of seven example dashboards in the docs, each with its ready-made prompt: asset-level data quality, alert operations, monitoring coverage, table health, credit consumption, data product scorecards and monitor results.
  • Change it by asking: "Only the tables the Risk domain owns." "Last 30 days, not 7." "Add median time to fixed."
  • Data quality by table: Score a domain, data product or tag by the tables in it, not by the monitors that cover them.
  • Trends over time: See how alerts, credits and asset health change over time, by day, week or month. Each metric's page in the docs shows how far back it goes.
  • Follows your access: Your AI tool sees the same domains and data you can see in Monte Carlo, nothing more.
  • The account-analytics dashboard skill: In coding agents (Claude Code, Cursor, Codex, GitHub Copilot), install the dashboard skill from the Agent Toolkit. Your dashboards follow the same layouts every time and link straight back to the tables, monitors and alerts in Monte Carlo.
  • Query it directly: Prefer to work in a script or notebook? Call the Analytics API yourself. The data catalog documents every metric and dimension.

For best results, use a capable model, such as Claude Sonnet or Opus, or GPT, together with the dashboard skill. If your AI tool doesn't support skills, point it to the dashboard examples and style pages instead.

Learn more & try it: Analytics API · How to query · Data catalog · Dashboard examples · Dashboard style · MCP server · Agent Toolkit

Availability: Public preview, on for every account. Monitor run metrics, such as validation monitor results, are in private preview. Contact your Monte Carlo account team to get access.