
The Universal Semantic Layer for Analytics and AI
Cube is an open-source agentic analytics platform built on a universal semantic layer that connects data sources to BI tools, embedded analytics, LLMs, and AI agents. It enables data teams to define metrics once and deliver consistent, governed data across every downstream tool via SQL, REST, and GraphQL APIs.
Define metrics, dimensions, and business logic once in a centralized data model consumed consistently by all downstream tools.
Natural language analytics chat powered by LLMs for conversational data queries via embedded interfaces.
Automatically computes and stores pre-aggregated query results to speed up response times and reduce warehouse costs.
Exposes data via SQL, REST, and GraphQL APIs so any tool or AI agent can consume it natively.
SDK and APIs to embed dashboards, workbooks, and analytics chat directly into customer-facing SaaS applications.
Syncs the semantic layer to BI tools like Tableau, Looker, and Power BI for consistent metric definitions.
Fine-grained data access policies applied at the semantic layer ensuring users only see authorized data.
Use Cube as the analytics backend for customer-facing dashboards with workbooks, charts, and AI-powered analytics chat.
Define all metrics upstream once and sync consistent definitions to Tableau, Looker, Power BI, and other tools.
Give AI agents governed, structured access to business data through Cube's APIs for accurate, context-aware answers.
Cache common query patterns with pre-aggregations to reduce expensive queries hitting cloud warehouses.
Interactive no-code workbooks for business users to explore data and build dashboards without writing SQL.
Business users explore data and build ad-hoc reports with Cube Workbooks without needing SQL knowledge.

Customer-facing analytics for any platform