Overview

The Create knowledge DB node is a super node that creates new knowledge databases from processed embeddings or segments using specialized knowledge database services. This node serves as a focused database creation tool for workflows where embeddings or segments have already been prepared and need to be organized into searchable knowledge databases.

The node provides access to multiple knowledge database service families to deliver flexible database creation capabilities:

• FAISS Vector DB services for high-performance local vector similarity search
• Pinecone services for cloud-managed vector databases with automatic scaling
• Elasticsearch services for distributed search with hybrid text and vector capabilities
• MongoDB services for document databases with integrated vector search

Unlike the Real-time knowledge injector which can handle the full pipeline from raw content to knowledge database, this node focuses specifically on the final database creation step from prepared data. It accepts either pre-generated embeddings for vector databases or segments for text-based databases, making it ideal for workflows with separate embedding generation or when using existing processed data.

Node Specification

Core Properties

Note: This is a super node that depends on selected services for its functionality. The parameters, input ports and output ports below are the base set provided by the node. Selected services will add their specific parameters and may modify available options.

Parameters

Service-specific parameters will be added dynamically based on the selected knowledge database service.

Input Ports

Note: At least one input must be connected. Some services require specific input types.

Output Ports