Gmail to Vector Embeddings with PGVector and Ollama

This n8n workflow automates the process of extracting email content from Gmail, converting it into vector embeddings using Ollama, and storing these embeddings in a PostgreSQL database with the PGVector extension. This allows for semantic search and AI-powered analysis of your email data.

What it does

This workflow simplifies and automates the following steps:

1. Triggers on new Gmail messages: It listens for new emails arriving in a specified Gmail account.
2. Filters out unwanted emails: It includes an "If" node, suggesting a conditional check to process only relevant emails. (The specific condition is not visible in the JSON but implies filtering logic).
3. Extracts and Prepares Email Content: The workflow processes the email body and subject.
4. Splits text into manageable chunks: It uses a "Recursive Character Text Splitter" to break down long email content into smaller, more digestible pieces suitable for embedding.
5. Generates Vector Embeddings: It leverages Ollama to create vector embeddings from the text chunks.
6. Stores Embeddings in PGVector: The generated vector embeddings, along with the original email content and metadata, are then stored in a PostgreSQL database configured with the PGVector extension.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• Gmail Account: An active Gmail account with appropriate permissions for n8n to access.
• Ollama Instance: A running Ollama instance accessible from your n8n environment, configured with the desired embedding model.
• PostgreSQL Database with PGVector: A PostgreSQL database with the pgvector extension installed and enabled. You'll need database credentials (host, port, database name, user, password).
• n8n Langchain Nodes: Ensure you have the @n8n/n8n-nodes-langchain package installed in your n8n instance to use the Text Splitter, Ollama Embeddings, and PGVector nodes.

Setup/Usage

1. Import the Workflow:
   • Download the provided JSON file.
   • In your n8n instance, click "Workflows" in the left sidebar.
   • Click "New" and then "Import from JSON".
   • Paste the workflow JSON or upload the file.
2. Configure Credentials:
   • Gmail Trigger: Configure your Gmail OAuth2 credentials.
   • Ollama Embeddings: Configure the connection to your Ollama instance.
   • Postgres PGVector Store: Configure your PostgreSQL database credentials.
3. Customize the "If" Node:
   • The "If" node (ID 20) is currently a placeholder. You will need to configure its conditions to filter emails based on your specific needs (e.g., sender, subject keywords, labels).
4. Activate the Workflow:
   • Once all credentials are set and the workflow is configured, activate it by toggling the "Active" switch in the top right corner of the workflow editor.

The workflow will now automatically process new emails according to your defined filters, generate embeddings, and store them in your PGVector database.