Skip to main content
A table is a dataset in the Catalog. Tables are created by Sources, Queries, Notebooks, and History, and every table lives inside a Layer, optionally organized into Folders. Under the hood, tables are stored as Delta Tables in your Lakehouse, so they are consistent, versioned, and ready for both analytics and processing.

​
The table overview

Open a table to see its details:
  • Its columns and their types.
  • Metadata such as size, total number of rows, and creation and last-update dates.
  • A link to the source or pipeline that generates it.
When a source is first configured, its tables appear in the Catalog before any run has filled them. Empty tables are expected at that point—trigger the source or wait for its schedule, and the data will land.

​
Name tables for humans and agents

Table names are part of the context your team and AI agents rely on. A person should understand what a table contains without opening it.
  • Use lowercase snake_case names.
  • Name the business entity and, when relevant, its grain: facebook_ads_campaign_daily is clearer than ads_data.
  • Keep source names when they help distinguish similar datasets: hubspot_deals and erp_orders.
  • Avoid unexplained abbreviations, temporary labels, and names such as table_1, new, or final.
  • Add descriptions to tables and important fields, especially when a business definition is not obvious.
  • External tables reference a BigQuery table in your own project and are read in place, without copying. They appear in the tables list with an External badge.
  • Volumes store non-tabular files such as PDFs, images, and videos alongside your tables.

Layers

The storage and permission boundary a table lives in.

Folders

Group tables inside a Layer.

Lineage

See how tables and pipelines depend on each other.

Expectations

Validate the data in your tables.