Cube

What is Cube?

Cube is an analytics platform built on a semantic layer that delivers governed business intelligence to both human users and AI agents. It serves data teams and product developers who need consistent metric definitions across dashboards, embedded applications, and agent-driven workflows. Access requires signing in or requesting a demo.

What are the features of Cube?

• Semantic model

  Defines metrics once so every downstream tool, dashboard, and AI agent uses the same calculation logic and governance rules.
• Analytics Chat

  Accepts natural-language questions and builds queries against the semantic model to return governed answers.
• Workbook Agent

  Assists analysts by drafting and iterating on SQL queries within a combined visual and code editor.
• Dashboard Agent

  Builds charts conversationally, supported by pre-aggregation and caching for sub-second performance.
• MCP server

  Connects Cube to external tools such as Claude, ChatGPT, Cursor, and custom agents for structured data requests.
• Embedded Analytics

  Provides multi-tenant iframes, core data APIs, and a creator mode for shipping analytics inside third-party products.

What are the use cases of Cube?

• Customer success self-service

  Customer success managers answer their own questions by chatting directly with governed data instead of requesting analyst reports.
• Embedded product analytics

  Software teams ship branded, multi-tenant dashboards and workbooks inside their application without rebuilding a BI stack.
• Agent-to-agent data retrieval

  Autonomous AI agents request structured analysis via MCP to inform decisions without waiting for human intervention.