LanceDB Node icon

LanceDB Node

Createdvurtnec

Enables vector similarity search using LanceDB and custom embedding functions within a Node.js environment.

About

This project provides a Node.js implementation for performing vector search operations using LanceDB, leveraging Ollama's embedding model for generating vector embeddings. It showcases how to connect to a LanceDB database, create custom embedding functions, perform vector similarity searches against stored documents, and process/display search results. The setup allows for integration with Claude Desktop as an MCP service, streamlining the process of incorporating vector search capabilities into various applications.

Key Features

  • Performs vector similarity search against stored documents
  • Connects to LanceDB for vector storage and retrieval
  • Includes example script for testing vector search functionality
  • Utilizes Ollama for custom text embedding generation
  • Integrates with Claude Desktop as an MCP service

Use Cases

  • Implementing semantic search in Node.js applications
  • Building retrieval-augmented generation (RAG) systems
  • Enhancing search capabilities with vector embeddings