Supabase Vector Database Integration

This uide walks you through setting up a Retrieval-Augmented Generation (RAG) pipeline in Vectorize using Supabase as your vector database.

What is Supabase?

Supabase is an open-source Firebase alternative that provides a PostgreSQL database with built-in vector search capabilities. It's a great choice for beginners because it offers:

• A user-friendly interface
• Built-in authentication
• Simplified database management
• PostgreSQL compatibility with the pgVector extension

Before you Begin

Ensure you have:

• A Vectorize account (Create one free here ↗)
• A Supabase account (Create one on Supabase ↗)
• A Supabase project with the pgVector extension enabled (see the Quickstart Guide for setup instructions)

Configure the integration

To connect Vectorize to your Supabase database:

1. Click Vector Databases from the main menu.

2. Click New Vector Database Integration from the Vector Databases page.

3. Select the Supabase card.

   

4. Enter the parameters in the form using the Supabase Parameters table below as a guide, then click Create Supabase Integration.

   

Supabase Parameters

Field	Description	Required	Where to find it
Name	A descriptive name to identify the integration within Vectorize.	Yes	Choose any name that helps you remember what this connection is for.
Host	The hostname of your Supabase PostgreSQL database.	Yes	You can find this in the connect tab under the Session Pooler section.
Port	The port number to connect to your database.	Yes	Usually 5432 for Supabase
Database	Your database name.	Yes	You can either choose and existing database or we will automatically create one for you if it doesn't exist.
Username	The username used to connect to your database.	Yes	You can find this in the connect tab under the Session Pooler section.
Password	The password for your database user.	Yes	This is the password you set when creating your Supabase project.

You can find the host, and username by selecting the connect button in your Supabase project. Scrolling down to Session pooler and selection 'View parameters'.

Configuring the Supabase integration in a RAG Pipeline

The Supabase integration has two parts:

1. Authentication: This connects to your Supabase database and can be reused across multiple pipelines.
2. Pipeline Configuration: This specifies which table in your Supabase database to use for storing vector embeddings.

When configuring your RAG Pipeline, you'll need to specify a table name. If the table doesn't exist yet, Vectorize will automatically create it for you with the proper schema for storing vector embeddings.

Understanding pgVector

Supabase uses the pgVector extension to store and search vector embeddings. This extension:

• Adds vector data types to PostgreSQL
• Enables similarity search operations
• Supports multiple indexing methods for efficient vector search

Vectorize automatically handles the technical details of working with pgVector, so you don't need to write any SQL queries or understand the underlying implementation.

What's next?

• If you haven't yet built a connector to your source database, go to configure a source connector and select your source data platform.

  OR

• If you're ready to start creating pipelines for vectorizing data, head to Pipeline Basics.