Skip to main content

Vector Store Dataset

Overview​

VectorStoreDataset is a dataset component responsible for supplying selected vector-store indexes to pipelines.

Its role is limited to scoping vector-store documents and passing them to the appropriate pipelines for execution.

Purpose​

  • Selects a specific vector-store dataset
  • Restricts access to configured vector-store indexes
  • Provides document references and metadata to pipelines
  • Ensures consistent document scope across agents and workflows

Configuration Contents​

The VectorStoreDataset configuration defines which vector-store data is exposed to pipelines.

  • Dataset – Unique dataset from which the selected vector documents are resolved.

  • Document Selection – List of selected vector-store index that restrict and secure pipeline access.

Responsibility Scope​

What VectorStoreDataset Does​

  • Fetches references to selected vector-store documents
  • Supplies dataset and document context to pipelines
  • Acts as a document provider, not a retriever
  • Maintains strict separation between data selection and retrieval logic

Info: VectorStoreDataset is used together with the following pipelines:

Summary​

VectorStoreDataset is a document-scoping dataset for vector-store–backed workflows.
It strictly controls which vector documents are exposed and delegates all retrieval, ranking, and semantic logic to pipelines.