Lancedb

🚀 LanceDB Node.js Vector Search

A sample project for vector search implemented in Node.js, integrating LanceDB and Ollama's embedding models.

🚀 Quick Start

This project demonstrates how to:

• Connect to a LanceDB database.
• Create a custom embedding function using Ollama.
• Perform vector similarity searches to find stored documents.
• Process and display search results.

✨ Features

• Connect to LanceDB and execute vector similarity searches.
• Use Ollama to create custom embedding functions.
• Integrate with MCP services in Claude Desktop.

📦 Installation

Prerequisites

• Node.js (v14 or higher).
• A locally running Ollama service configured with the nomic-embed-text model.
• Permissions to read from and write to the LanceDB storage location.

Steps

1. Clone the repository.
2. Install dependencies:

pnpm install

📚 Documentation

Dependencies

• @lancedb/lancedb: A LanceDB client for connecting with Node.js.
• apache-arrow: For handling columnar data.
• node-fetch: For sending requests to the Ollama API.

Usage

Run the vector search test script:

pnpm test-vector-search

Or execute directly:

node test-vector-search.js

Configuration

The script connects to the following services:

• LanceDB at the specified path.
• Ollama API at http://localhost:11434/api/embeddings.

MCP Configuration

To use it as an MCP service in Claude Desktop, add the following to the MCP configuration JSON file:

{
  "mcpServers": {
    "lanceDB": {
      "command": "node",
      "args": [
        "/path/to/lancedb-node/dist/index.js",
        "--db-path",
        "/path/to/your/lancedb/storage"
      ]
    }
  }
}

Replace the following paths with the actual ones:

• /path/to/lancedb-node/dist/index.js - The actual installation location of the Node.js service.
• /path/to/your/lancedb/storage - The actual path of the LanceDB storage directory.

Custom Embedding Function

The project includes a custom OllamaEmbeddingFunction that is responsible for:

• Sending text to the Ollama API for processing.
• Receiving 768-dimensional embedding vectors.
• Formatting the output to meet the requirements of LanceDB.

Vector Search Example

The example code demonstrates how to search for "how to define success criteria" in the "ai-rag" table and display the matching results and their similarity scores.

💻 Usage Examples

Basic Usage

pnpm test-vector-search

Advanced Usage

node test-vector-search.js

📄 License

MIT License

🔧 Contribution Guidelines

Any contributions are welcome! Feel free to submit a Pull Request.