Skip to content

01 Email - AI-Powered Email Response System

An automated workflow that monitors Gmail for unread emails, generates personalized AI responses using OpenAI's GPT-4, and stores the results in Airtable for review and management.

Purpose

No business context provided yet — add a context.md to enrich this documentation.

This workflow appears to be designed for automated email management, specifically for someone who receives many inquiries and needs AI assistance to draft appropriate responses while maintaining a personal touch and consistent messaging aligned with their brand and values.

How It Works

  1. Scheduled Monitoring: Every 2 hours, the workflow checks Gmail for unread emails
  2. Email Retrieval: Fetches all unread emails and their detailed content including attachments
  3. AI Response Generation: For each email, an AI agent analyzes the content and generates a personalized response based on predefined user context (Drake Surach from AiF company)
  4. Data Storage: Stores the original email details and AI-generated response in an Airtable database
  5. Email Management: Marks the processed emails as read to prevent reprocessing

The AI agent is specifically trained to respond as Drake Surach, a 21-year-old focused on teaching AI to his community, with clear guidelines about declining external opportunities while being helpful and professional.

Workflow Diagram

graph TD
    A[Schedule Trigger<br/>Every 2 Hours] --> B[Gets ALL Emails<br/>Unread Only]
    B --> C[Pulls GRANULAR Details<br/>Get Message Content]
    C --> D[AI Agent<br/>Generate Response]
    E[OpenAI Chat Model<br/>GPT-4] --> D
    D --> F[Creates Record in Airtable<br/>Store Email & Response]
    F --> G[Marks Emails Unread<br/>Mark as Read]

Trigger

Schedule Trigger: Runs automatically every 2 hours to check for new unread emails.

Nodes Used

Node Type Purpose
Schedule Trigger Initiates workflow every 2 hours
Gmail (Gets ALL Emails) Retrieves unread emails from Gmail
Gmail (Pulls GRANULAR Details) Fetches detailed email content and attachments
AI Agent Generates personalized email responses using predefined context
OpenAI Chat Model Provides GPT-4 language model for AI agent
Airtable (Creates Record) Stores email data and AI responses
Gmail (Marks Emails Unread) Marks processed emails as read
Structured Output Parser Formats AI responses (connected but not actively used)

External Services & Credentials Required

Gmail

  • OAuth2 credentials for Gmail API access
  • Scopes needed: Read emails, modify email labels/status
  • Account: Must be the Gmail account to monitor

OpenAI

  • API Key for GPT-4 access
  • Model: gpt-4o configured in the Chat Model node

Airtable

  • API Key or Personal Access Token
  • Base ID: app4qQr7DGv3at14a (Email Collection System)
  • Table ID: tblmUrFXNdzDCve4a (Email Aggregation)

Environment Variables

No environment variables are explicitly used in this workflow. All configuration is done through node parameters and credential connections.

Data Flow

Input

  • Unread Gmail messages with subject, body, sender information, and attachments

Processing

  • Email content is analyzed by AI agent with context about Drake Surach's persona and business focus
  • AI generates appropriate responses based on email type (inquiries, opportunities, community interest)

Output

  • Airtable Record containing:
    • Sender Name
    • From Email
    • Subject
    • Email Body
    • AI-Generated Draft Reply
    • Thread ID

Error Handling

The workflow does not include explicit error handling nodes. Potential failure points: - Gmail API rate limits or authentication issues - OpenAI API failures or quota exceeded - Airtable API connection problems - Email parsing errors for malformed messages

Consider adding error handling nodes and notification systems for production use.

Known Limitations

  • No explicit error handling or retry mechanisms
  • Processes all unread emails without filtering (could include spam or irrelevant emails)
  • AI responses are stored but not automatically sent (requires manual review and sending)
  • No duplicate detection if workflow runs multiple times on same emails
  • Limited to Gmail only (no other email providers)

No related workflows mentioned in the provided context.

Setup Instructions

  1. Import Workflow

    • Import the JSON into your n8n instance
    • The workflow will be inactive by default

  2. Configure Gmail Credentials

    • Set up Gmail OAuth2 credentials in n8n
    • Test connection with your Gmail account
    • Ensure proper scopes for reading and modifying emails

  3. Configure OpenAI Credentials

    • Add your OpenAI API key to n8n credentials
    • Verify GPT-4 access in your OpenAI account

  4. Configure Airtable

    • Set up Airtable credentials (API key or PAT)
    • Verify access to the specified base and table
    • Ensure table schema matches the expected fields

  5. Customize AI Agent

    • Review the system message in the AI Agent node
    • Update user information (name, company, age, aspirations) as needed
    • Modify examples to match your communication style

  6. Test the Workflow

    • Run manually first to verify all connections work
    • Check that emails are properly retrieved and processed
    • Verify Airtable records are created correctly

  7. Activate

    • Enable the workflow to start automatic processing
    • Monitor initial runs to ensure proper operation

  8. Optional Enhancements

    • Add error handling nodes
    • Implement email filtering logic
    • Set up notifications for failures
    • Add automatic email sending (with appropriate safeguards)