Input node
Salesforce input
Read Salesforce objects with SOQL, paged and typed automatically.
What it does
Runs a SOQL query against Salesforce through a saved Connection and streams the records into the pipeline, following Salesforce's own paging to the end. Either pick an object and list the fields you want, or write the whole query yourself when you need a filter, a relationship or an aggregate the two boxes can't express.
Inputs and outputs
| Inputs | None. This node starts a pipeline. |
| Outputs | One. |
When to use it
- Extract accounts, contacts or opportunities for a warehouse load
- Report across Salesforce and another system in one pipeline
- Sync a filtered slice of an object on a schedule
Examples
Whole object
Object: Account
Reads every standard field: fine to start with, wide in production.
Chosen fields
Object: Contact · Fields: Id, Email, CreatedDate
Smaller and faster, and the columns stop moving when someone adds a field.
Your own query
SELECT Id, Name FROM Account WHERE CreatedDate = LAST_N_DAYS:7
A written query wins over the object and field boxes.
A custom object
Object: My_Thing__c
Custom objects read exactly like standard ones, __c suffix included.
Tips
- A written SOQL query wins over the object and field boxes, so clear it to go back to using them.
- The token is minted per run and refreshed if it expires mid-read, so long extracts don't die at the hour mark.
- Use “Fetch sample” to fill in the columns, and the builder shows them with no run.
- SOQL is not SQL: there is no SELECT *, and you have to name every field you want.
- Max records is a real cap: it stops mid-page rather than delivering the rest of the page you were on.