1. API Reference
  2. Push Data



name of Sieve project


name to assign to the image or video source


direct URL to data source

This endpoint allows you to push a data source to Sieve. Conceptually, a data source is a single image or video you’d like to process through a given project’s defined workflow. You can give it a unique name which you use to access it later. Or you can access it by the job_id that’s returned to you.

Sieve ingests images and videos in the form of URLs, which are either public or signed (private and secure).

Learn more about signed URLs for AWS or GCP at the respective links. In addition, you need to supply a unique video name for each video. Optionally, you can also submit custom metadata for the video as a whole, along with its associated value. For example, if you have internal indexes or fields you want to track with our platform, this would be an easy way for you to access those within Sieve as well.

Sieve is also experimentally testing streaming video ingestion. If you would like to test this feature, please reach out to us!