The agentic analytics platform built on a semantic layer
Cube gives humans and AI agents governed, reliable access to trusted business data. People work in Analytics Chat, workbooks, dashboards, and embedded experiences; Claude, Codex, and other agents use the same business context through MCP, APIs, or CLI.
and 50+One governed BI platform for people and agents
BI for agents means AI agents can use the analytics platform as first-class users — without a parallel path to raw data. Cube gives people and agents the same metrics, permissions, and analytical context across internal and embedded analytics.
- Model the business onceMetrics, dimensions, entities, joins, and access rules give every person and agent the same definition of the business.
- Explore through governed workflowsAgents discover approved concepts, query them under the asker's permissions, and iterate without re-deriving SQL from raw tables.
- Create work people can inspectAnswers, reports, workbooks, dashboards, and model changes retain the context and lineage people need to review the result.
The AI analyst your whole team can talk to
- Query in natural languageAsk in plain English. The agent writes and runs the queries against your semantic model — no SQL required.
- Ask follow-up questionsRefine iteratively; context, filters, and prior queries carry forward across the thread.
- Answers you can trustCharts, tables, and the reasoning behind them — every number governed and scoped to each user.
An AI analyst inside every workbook
- Build with the Workbook AgentReports, charts, and dashboards from a request — while you keep control of every field, filter, and label.
- Three ways to queryPoint-and-click, raw SQL, or Semantic SQL on top of governed metrics — all in one place.
- Governed by constructionEvery chart is backed by a reusable, governed report.
Dashboards that answer the follow-up
- Drill deeper with the Dashboard AgentBreak a metric down, extend the range, or explain a definition — read-only, on the same governed model.
- Every widget you needChart, KPI, filter, time-grain, text, and AI Summary widgets.
- Scheduled and embeddableScheduled refreshes, PNG and PDF snapshots, and embed-anywhere.
AI that builds your model — and runs on it
- Bootstrap with the Semantic Model AgentGenerate a model from your warehouse tables, convert existing LookML, or extend definitions in SQL-first YAML.
- Grounded, not guessingEvery agent reasons over governed definitions instead of raw tables — which is what makes the answers trustworthy.
- Governed end to endRow-level security flows from the model to every answer, on any warehouse.
Give Claude, Codex, and every agent governed business context
Connect the AI tools your team already uses through MCP, APIs, or CLI. Each agent reaches the same semantic model and production controls as every Cube surface.
- Use the agent your team already trustsClaude, Codex, ChatGPT, Cursor, or one you build can work with governed business data instead of a separate copy of the logic.
- Analytics in SlackDrop the Slack Agent into a thread; ask and follow up where your team already works.
- Permissions carry throughCan't see a metric in Cube? You won't see it in Slack either.
Frequently Asked Questions
Teams building agentic analytics on Cube
“The future of reporting isn't a chart, it's an insight. Large language models are becoming a commodity — the LLM is the engine, but the semantic layer is the map. A well-modeled ontology is the difference between 'I don't understand that question' and a correct, contextualized answer with a chart and a clear explanation. Cube gives us the foundation to make that real for every customer.”
Read the story“Cube becomes our single source of truth for metric definitions and powers everything from customer-facing dashboards to AI-driven quarterly business reviews. CSMs gain back dozens of hours each quarter, enabled by Cube’s semantic layer and agentic analytics.”
“We integrated Cube Cloud smoothly with ClickHouse, leveraging both for fast query execution while maintaining the abstraction needed for different teams to access data without diving into database-specific complexities.”
“Without Cube, our data analysts might have to write 20 different queries for a single core business metric. With Cube, that metric is defined once in the data model, and every downstream tool uses that definition along with the associated calculation logic.”