Agentic analytics on top of Databricks, built on a semantic layer
Cube is the agentic analytics platform that sits on top of Databricks. It turns your Databricks data into governed metrics that every dashboard, spreadsheet, embedded app, and AI agent reads the same way — so business users and agents get answers they can trust. Connect Databricks, model your metrics once, and serve them everywhere in minutes.
Answers people trust
business users and AI agents query Databricks in plain language and get metrics your data team governs.
Governance that travels
row-, column-, and user-level rules modeled once, enforced across every tool and API.
Faster queries, lower cost
high-performance caching on top of Databricks keeps responses sub-second and cuts warehouse load.
Make Your Databricks Data AI- and BI-Ready
Databricks powers your storage and compute, but teams still redefine metrics in every BI tool and spreadsheet — so the numbers disagree. Cube sits on top of Databricks as the governed layer every consumer reads from: model "revenue" or "active user" once, and dashboards, spreadsheets, embedded apps, and AI agents all return the same number.
Deliver Your Databricks Data Where It's Needed
Serve your governed Databricks metrics to every consumer from one model — a Postgres-compatible SQL API for BI, REST and GraphQL for embedded analytics, MDX for Excel, and DAX for a native Power BI connection. No re-implementing logic in each tool, and no metric sprawl.
Secure Your Databricks Data at the Semantic Layer
Define row-level, column-level, and user-level access rules upstream, in the semantic layer, and they apply automatically across every tool and API — even into AI agents and spreadsheets. Governance follows the data, so you stay compliant without locking people out of the tools they prefer.
Optimize Databricks Performance and Reduce Compute Costs
Cube's high-performance caching sits on top of Databricks, delivering sub-second API responses under high concurrency while cutting warehouse load and cost. It's a performance layer for every connection, from dashboards to agents.
Ground AI Agents on Trusted Context from Cube
Cube's AI agents work natively on top of Databricks, grounded in your governed semantic model. They automate analysis, explain their outputs, and act within your governance guardrails — Databricks is the source, and the semantic layer is what makes their answers trustworthy.
“I see Cube as a key technology in our data stack because it serves a very real purpose in adding semantics, speed, and security administration at the API layer. It's something that you don't want to build yourself. I can see us using Cube much more in the future.”Read the story