How to create an ETL pipeline

Build a pipeline that reads data from a source, transforms it, and writes to a destination.

~25 min | Role: Weave Designer

Before you begin

You need the Weave Designer role and at least one datasource configured. Your pipeline will read from and write to datasources.

From the Weave navigation, click Design → New workflow.

From the node palette, drag a DB input, file input, or API input onto the canvas. Connect it to the datasource.

Chain filter, join, aggregate, map, split, or lookup nodes to shape the data as needed.

Drag a DB write, file export, or API push node. Connect it to the target datasource.

Draw connections between nodes to define flow direction. Data flows left to right.

Click Test to run with sample data. Check the execution log for errors.

Save the workflow, then optionally schedule it to run on a cadence.

For logic not covered by built-in transforms, use a custom script node — it lets you write arbitrary transformation code inline.