Couchbase Vector Store icon

Couchbase Vector Store

Couchbase Vector Store Node

Couchbase Vector Store

AI-generated

Overview

This node integrates with a Couchbase Vector Store to perform various operations related to vector-based document storage and retrieval. It supports inserting documents, retrieving ranked documents based on similarity queries, updating documents by ID, and retrieving documents for use as vector stores or tools in AI workflows. This node is useful in scenarios where users want to manage and query vector embeddings for AI applications such as semantic search, recommendation systems, or AI agent tools.

Use Case Examples

  1. Insert multiple documents with vector embeddings into a Couchbase vector store for later similarity search.
  2. Retrieve the top ranked documents from the vector store based on a search prompt to feed into an AI chain or tool.
  3. Update existing documents in the vector store by their IDs to keep embeddings and metadata current.
  4. Retrieve documents as a tool with a description for use by AI agents to perform specific tasks.

Properties

Name Meaning
Operation Mode Selects the operation to perform on the Couchbase vector store, such as inserting documents, retrieving documents, or updating documents.
Bucket The Couchbase bucket to use for the operation, selectable from a list or by name.
Scope The Couchbase scope within the bucket to use for the operation, selectable from a list or by name.
Collection The Couchbase collection within the scope to use for the operation, selectable from a list or by name.
Index The name of the search index to use for the operation.
Embedding Batch Size Number of documents to embed in a single batch (used when inserting documents).
Prompt Search prompt to retrieve matching documents from the vector store using similarity-based ranking (used when retrieving documents).
Limit Number of top results to fetch from the vector store (used when retrieving documents).
Include Metadata Whether to include document metadata in the results (used when retrieving documents).
ID ID of an embedding entry to update (used when updating documents).
Description Description explaining what the tool does, to help LLMs produce expected results (used when retrieving documents as a tool).

Output

JSON

  • documents - Array of documents retrieved or affected by the operation, including their vector embeddings and optionally metadata.
  • tool - When retrieving as a tool, the output includes a tool object usable by AI agents.

Dependencies

  • Requires a Couchbase API credential for authentication and access to the Couchbase cluster.

Troubleshooting

  • Common errors include unsupported operation modes if the mode parameter is set incorrectly. The node only supports 'insert', 'load', and 'update' modes in execute(), and 'retrieve' and 'retrieve-as-tool' modes in supplyData().
  • Errors may occur if the specified bucket, scope, or collection does not exist or is inaccessible. Verify resource names and permissions.
  • Batch size for embedding insertion should be set appropriately to avoid performance issues or timeouts.

Discussion