Deliver to a destination
Connect Google Drive, bind it to a pipeline's Destination stage, and confirm delivery — while every result stays available over the Web Response API.
A pipeline always returns its result over the Web Response API. Connecting Google Drive adds a second delivery: the finished output is written to a Drive folder when the job completes. This guide connects Drive and binds it to a pipeline.
Workspace admins and pipeline builders. You need a Google account with access to the input and output folders you want to use.
1. Connect Google Drive
Go to Settings → Integrations → Connectors and choose Add connector. The wizard runs in four steps — Name, Connect, Folders, Done.
Step one of the Connect Google Drive wizard. The name is shown on the Integrations page and in the pipeline picker, so make it recognisable.
At Connect, sign in to Google and grant access. At Folders, pick the input folder the connector watches and the output folder it writes results to. Done saves the connector; it appears on the Connectors tab and begins polling its input folder.
2. Bind the connector to your pipeline
Open the pipeline in Design and select the Destination stage. Under Output destination, add the connector as a Deliver to Google Drive target. Results for this pipeline are now written to the connector's output folder.
The Destination inspector. Web Response is always on; the Google Drive target is what you add here. Routing lives on this stage — the output format does not.
3. Keep the Web Response
Leave Web Response on. It cannot be turned off, and that is deliberate: even with Drive delivery configured, every result stays retrievable over the Job Results API and in the API Playground, so an integration is never blocked waiting on a folder sync.
4. Set the output format
The Destination stage routes the result; it does not shape it. Set the output format on the Post Processing stage inspector — that is where a pipeline decides, for example, that it emits JSON. Confirm the format there before you rely on what lands in the output folder.
5. Confirm the delivery
Run a production job through the pipeline and check the connector's output folder for the result. A test run is the wrong tool here — test runs return their result over Web Response but never write to a connector — so use a real submission, or see Test a document in a pipeline for the inspection-only path.
Idempotency and retry storms
A connector feeds one pipeline and writes each job's result once, keyed to the job, so a re-poll or a retry does not duplicate output. If you need the result elsewhere the moment it is ready — rather than after the folder write — subscribe to it with a webhook instead of watching the folder.