Skip to main content

Pipelines for Snowflakes

Pipelines in Snowflake are built from streams, tasks and dynamic tables. A stream records row changes on a table, a task runs SQL on a schedule or when triggered, and a dynamic table keeps a query result up to date from a declared target lag. Together they support incremental transformation without an external scheduler.

Stage: Transformation · Part of the Snowflake roadmap · Last reviewed 19 November 2026

What to learn in Pipelines

Pipelines interview questions

What does a Snowflake stream store?

A stream stores an offset into a table's change history, not a copy of the data. Querying it returns the rows changed since that offset, with metadata columns showing the action. The offset advances when a DML statement consumes the stream in a transaction.

How does a dynamic table differ from a materialised view?

A dynamic table is defined by a query and refreshed to meet a target lag, and it can use joins and complex logic that Snowflake manages incrementally where it can. Materialised views have tighter query restrictions and refresh automatically on a single base table.

Where Pipelines fits in the Snowflake roadmap

Pipelines belongs to the Transformation stage (intermediate level). Build pipelines.

Back to the full Snowflake roadmap