Embedding

Embedding Processor #

Generates vector embeddings for document chunks using AI models.

This processor enables semantic search and retrieval by converting text chunks into dense vector representations.

Configuration #

ParameterTypeRequiredDefaultDescription
message_fieldstringdocumentsThe field in the pipeline context containing the documents to process
output_queueobjectnullOptional queue configuration for sending processed documents to a output queue
model_providerstringYes-ID of the AI model provider for embeddings
modelstringYes-Name of the embedding model (e.g., text-embedding-3-small)
embedding_dimensionint32Yes-Vector dimension (must match model’s supported dimensions)

Example #

- document_embedding:
  model_provider: openai
  model: text-embedding-3-small
  embedding_dimension: 1536
Edit Edit this page