Automated Multilingual Gmail Draft Reply with OpenAI GPT-4o

This n8n workflow automates the drafting of multilingual email replies in Gmail using OpenAI's GPT-4o model. It listens for new emails, processes their content, and generates a suitable draft reply, providing a significant boost to email productivity and cross-language communication.

What it does

1. Monitors Gmail: Continuously checks for new emails in your specified Gmail account.
2. Conditional Logic: Evaluates incoming emails based on defined conditions (though the specific conditions are not detailed in the provided JSON, it indicates a branching logic).
3. Routes Email Content: Directs the email content based on the initial conditional evaluation.
4. Generates Reply with AI: Utilizes an OpenAI Chat Model (GPT-4o is indicated by the directory name, though the JSON only specifies a generic "OpenAI Chat Model") to generate a draft reply.
5. Processes AI Output: Structures and parses the AI-generated response using a Structured Output Parser.
6. Creates Gmail Draft: Composes and saves the AI-generated reply as a draft in Gmail, ready for review and sending.

Prerequisites/Requirements

• n8n Instance: A running n8n instance (cloud or self-hosted).
• Gmail Account: A Google account with Gmail enabled.
• Google OAuth2 Credentials: Configured Google OAuth2 credentials in n8n for Gmail access.
• OpenAI API Key: An OpenAI API key with access to the GPT-4o model (or the model configured in the "OpenAI Chat Model" node).
• Langchain Integration: Ensure the @n8n/n8n-nodes-langchain package is installed in your n8n instance, as it is used for the LLM Chain and OpenAI Chat Model nodes.

Setup/Usage

1. Import the Workflow: Download the provided JSON and import it into your n8n instance.
2. Configure Gmail Trigger:
   • Open the "Gmail Trigger" node.
   • Add your Gmail OAuth2 credentials.
   • Configure the trigger to listen for new emails (e.g., specify labels, unread status, etc.).
3. Configure OpenAI Chat Model:
   • Open the "OpenAI Chat Model" node.
   • Add your OpenAI API key credentials.
   • Ensure the correct model (e.g., gpt-4o) is selected if not already configured.
   • Review and adjust the prompt within the "Basic LLM Chain" node to guide the AI in generating appropriate replies.
4. Configure Gmail (Draft) Node:
   • Open the "Gmail" node (responsible for creating drafts).
   • Add your Gmail OAuth2 credentials (the same as the trigger).
   • Map the output from the "Structured Output Parser" to the email draft fields (recipient, subject, body).
5. Review Conditional Logic (If/Switch Nodes):
   • Examine the "If" and "Switch" nodes. These nodes are placeholders for specific routing logic. You will need to define the conditions based on your specific requirements (e.g., filter emails by sender, subject keywords, sentiment, etc.) to determine which emails should trigger an AI-generated reply.
6. Activate the Workflow: Once all credentials and configurations are set, activate the workflow.

The workflow will now automatically monitor your Gmail inbox and draft replies based on your defined logic and the capabilities of the OpenAI model. Remember to review the drafted emails before sending them.