Introduction to the Cassis data intelligence layer

Cassis reads your data environment, builds a structured layer of business meaning on top of your data (the ontology), and lets anyone or any agent ask business questions in natural language. Every answer is grounded in the ontology, comes with the SQL that produced it, and carries full provenance.

The ontology sharpens with every use, as every correction and clarification from chat feeds back into the definitions.

You can use Cassis in two ways: through the web app, where you chat with your data and curate the ontology, or through the MCP server, which exposes the same engine to AI agents in Claude Code, Claude Desktop, Cursor, and any other MCP client.

Access is limited to selected design partners today. Leave your email at getcassis.com to be notified when it opens up.

What is an ontology?

An ontology is a structured representation of how your business thinks about its data: entities, relationships, metrics, dimensions, and the business logic and context behind each one. It sits between your raw warehouse tables and whoever is asking questions.

It's not a metrics file or a data catalog. Those define individual calculations or document what exists. An ontology captures what things mean in your specific context, how they connect, and their full provenance: origin, authorship, and edit history for every definition.

How it works

Reads your environment
Connects to your dbt project, warehouse schema, and documentation. Combines all of them to infer definitions.
Builds a structured ontology
Objects, properties, links, and business context. Each carries the source it came from so you can trace a definition back to a dbt model, a SQL column, or a piece of docs.
Tracks provenance
Every definition records where it came from. An edit history captures who changed what, when, and why.
Improves through use
Every question asked, every ambiguity clarified, every correction made feeds back. The ontology gets sharper as a side effect of normal work.
Serves humans and agents
The same engine powers the web app and the MCP server. Humans and AI agents query the same ontology, with the same grounding, and get the same auditable answers.

Pick your interface

Cassis ships two interfaces on top of the same engine. Most teams use both.

For humans
Using the web app
Chat with your data in natural language, inspect the proposed query plan before running it, and curate the ontology as you go. Five sections: Chat, Ontology, Suggestions, Evals, Tables.
For agents
Using the MCP server
A hosted MCP server that exposes the same grounded query engine. Connect from Claude Code, Claude Desktop, Cursor, or any custom agent. Three tools today: ping, list_projects, ask_question.

Interested?

We're building the data intelligence layer. Sign up to hear from us first.

Stay tuned