generated: '2026-07-21'
method: searched
source: https://docs.tecton.ai/docs/introduction
note: >-
  Derived from Tecton's documented object model (Concepts). Tecton has no public OpenAPI
  spec, so relationships are grounded in the framework's canonical entities rather than a
  machine-readable schema. Entities are the objects a user declares in a feature repository
  and manages via the SDK/CLI.
entities:
  - name: Workspace
    description: Top-level container for feature objects; separates development from live/production.
  - name: DataSource
    description: Raw data input (Batch, Stream, or Push) that feeds feature transformations.
  - name: Entity
    description: Business object a feature describes; defines the join keys features are keyed on.
  - name: Transformation
    description: Reusable transformation logic applied to data sources to compute features.
  - name: FeatureView
    description: >-
      Definition of one or more features (Batch, Stream, or Realtime), including the
      transformation, source data, and the entity it is keyed on. Produces feature values.
  - name: FeatureService
    description: >-
      A named group of features (from one or more FeatureViews) assembled for a specific
      model; the unit that is served online via the HTTP API and used to build training data.
  - name: Feature
    description: An individual measurable property produced by a FeatureView.
relationships:
  - from: Workspace
    type: has_many
    to: FeatureView
  - from: Workspace
    type: has_many
    to: FeatureService
  - from: Workspace
    type: has_many
    to: Entity
  - from: Workspace
    type: has_many
    to: DataSource
  - from: FeatureService
    type: has_many
    to: FeatureView
    via: features
  - from: FeatureView
    type: belongs_to
    to: Entity
    via: entities
  - from: FeatureView
    type: belongs_to
    to: DataSource
    via: source
  - from: FeatureView
    type: has_many
    to: Feature
  - from: FeatureView
    type: uses
    to: Transformation