Skip to main content
Orchestration components are used in orchestration pipelines, which handle the loading of data from source systems to a target data warehouse. Each component can have three output paths:
  • Green: continue if this component succeeded.
  • Gray: continue unconditionally, regardless of success or failure.
  • Red: continue if this component failed.
This allows you to define branching logic based on pipeline outcomes.

Component types

Connectors

Connectors pull data from external sources into your data warehouse. Examples include S3 Load, Salesforce Query, Excel Query, and Google Sheets.

Flow logic

Flow logic components control the sequence and conditions of pipeline execution. For example, the And component waits for multiple upstream components to succeed before proceeding.

DDL

DDL components manage data objects in your warehouse, such as tables, views, schemas, and streams. Examples include:
  • Create Table: creates or replaces a table.
  • Alter Table: modifies the structure of an existing table.
  • Truncate Table: removes all rows from a table without dropping it.
  • Schema Copy: copies a schema and its objects to a new location.
  • Zero Copy Clone: creates an instant clone of a table, schema, or database without duplicating the underlying data. Snowflake only.

Scripting

Scripting components allow custom logic, such as Python or SQL scripts.

AI

AI components let you integrate large language models (LLMs) directly into your orchestration pipelines:
  • Cortex Finetune—fine-tune an LLM using Snowflake Cortex on your own labeled dataset, then call the resulting model via the CORTEX.COMPLETE function in Snowflake.
  • Google Vertex AI Prompt—send prompts to a Google Gemini model using text, image, audio, or video inputs, and store the results in your data warehouse.

Run Transformation

Run Transformation is a special orchestration component that triggers a transformation pipeline as part of the orchestration flow, allowing you to embed transformation logic within an orchestration pipeline.
Each component is exclusively applicable to one pipeline type—orchestration or transformation—and cannot be added to the other. The one exception is the Run Transformation component.