Data Model#

There are only three core concepts you need to know.

  1. Files are the top level object for storing data. A File can store raw data and an ordered list of Blocks.

  2. Blocks are chunks of content within a File. They can contain raw data and/or text, and an unordered set of Tags.

  3. Tags are typed annotations on a Block or File.

The following diagram shows how data is created and used within Steamship:

Lifecycle of a File
  1. File raw data can either be created directly or imported via a File Importer

  2. Blocks on Files can either be created directly or created from raw data by a Blockifier

  3. Once you have blocks, you can run Generators and Taggers

  4. Find data that you need by querying

  5. Index data for search with the embedding search index