Semantic Catalog

Coming soon

Define metrics once—use them everywhere, by humans and AI alike.

Shared definitions, consistent answers

Create semantic layers per connection with namespaces like finance.revenue. Work with drafts, publish when ready, and keep every notebook and conversation grounded in the same logic.

YAML-based definitions with validation.
Drafts and versioning with explicit publish workflow.
Agents and notebooks resolve the same metrics automatically.

Semantic >campaigns.yaml

Published

1- id: campaigns

2type: model

3name: Campaigns

4table: marketing.campaigns

5dimensions:

6- id: channel

7name: Channel

8type: categorical

9- id: campaign_name

10name: Campaign Name

11type: categorical

12measures:

13- id: total_spend

14expr: SUM(spend)

1 model·2 measures·5 dimensions

Consistency

When everyone uses the same definitions, dashboards stop drifting and metrics become reusable building blocks instead of one-off SQL.

Shared definitions reduce dashboard drift.
Namespaces make it clear what's official versus ad-hoc.
Changes go through drafts before they affect production.

Velocity

Stop rewriting the same aggregation logic. Semantic layers let agents and humans reuse definitions—so teams move faster with less rework.

Agents and notebooks resolve the same metrics.
Teams iterate safely via drafts and publishing.
Less rework when stakeholders ask the same questions weekly.

Self-serve that doesn't drift

A semantic catalog makes shared context explicit—so answers are consistent across notebooks, conversations, and teams.