I Built A Custom AI Email Assistant with ChatGPT! | Make Automations 101 Tutorial

Introduction

In today's fast-paced digital world, managing your email can quickly become overwhelming. As a small business owner, I found myself inundated with numerous emails daily, especially since my Notion template business took off. To tackle this challenge, I decided to create a custom AI email assistant for my Gmail account using Make and ChatGPT. In this guide, I will walk you through the process of building your own email automation tool, designed to streamline responses and enhance productivity.

The Automation Journey Begins

After receiving countless inquiries about Notion templates, icon packs, and Discord community questions, I realized it was essential to develop a system that could help sort and respond to these emails. The main goals were:

1. Identify various types of emails and how to process responses.
2. Train a custom ChatGPT assistant to classify emails and eventually respond on my behalf.
3. Share all new emails with ChatGPT for analysis and processing.

To achieve these goals, I utilized Make, a no-code automation platform that integrates seamlessly with Gmail and other applications. The aim was to leverage AI intelligence from ChatGPT, creating a tailored assistant to help manage my specific business needs.

Understanding Make

Make is a powerful no-code automation tool that allows users to create workflows across various apps using a visual interface. It's easy to use, yet offers flexible and potent features for those who need advanced capabilities. Here's how we set up our integration:

1. Setting Up the Workflow: Within Make, I created a scenario that connects my Gmail account to my ChatGPT assistant. This integration would enable the AI to receive new emails, analyze their content, and categorize them automatically.
2. Flowchart Planning: To visualize the process, I used FigJam to create a flowchart detailing the different types of emails I receive and how the assistant would respond.
3. Email Classification: Based on my detailed flowchart, I categorized my emails into eight distinct types (e.g., partnership offers, discount requests, delivery issues) to aid in classification.

Building the Automation

To create the automation, I followed these steps:

1. Watch for New Emails: In Make, I added a Gmail module to watch for incoming emails. I set the criteria to filter out replies, ensuring that only new emails triggered the automation.
2. ChatGPT Integration: I added a ChatGPT module to process the incoming email's subject and body. This interface is vital as it transforms user queries into a format our AI can classify.
3. Routing the Output: Using routers within Make, I set up actions based on the classifications made by ChatGPT. Each category would trigger a specific response via Gmail to streamline communication with customers.

Creating the Custom ChatGPT Assistant

To build an effective AI assistant, I had to define clear parameters within ChatGPT. Here’s how I did it:

1. Setting Up a New Project: I created a custom GPT focused on email classification. I specified the task and the categories we'd defined in the flowchart.
2. Input and Output Definitions: I ensured the inputs (subject and email body) corresponded directly to the assistant's outputs (classification category).
3. Testing and Refining: The assistant underwent rigorous testing, where I provided example emails along with the expected output. This helped to fine-tune its response accuracy.

Results and Future Goals

Once the automation was fully operational, the AI assistant began labeling emails effectively, streamlining my response process and freeing up valuable time. The next phase of this project is to enhance the system further, allowing the assistant not only to classify emails but also to draft tailored responses based on previous communications.

The beauty of this setup lies in its simplicity and efficiency, showcasing how no-code solutions and AI technology can revolutionize email management.

Keyword

• AI email assistant
• Make
• ChatGPT
• Email automation
• Workflow
• No-code tool
• SMTP
• Email classification
• Gmail integration
• Productivity

FAQ

Q1: What is Make, and how does it help with email automation?
A1: Make is a no-code automation platform that allows users to connect various applications, making it easier to create workflows that can automate tasks, including email management.

Q2: Can I customize my ChatGPT assistant to fit my business needs?
A2: Yes! You can train your ChatGPT assistant with specific instructions and examples relevant to your business to improve its accuracy in handling inquiries.

Q3: Is it necessary to have coding experience to use Make?
A3: No, Make is designed for users without coding expertise, allowing anyone to create automations using a visual interface.

Q4: How does the automation classify emails?
A4: The automation uses a trained ChatGPT model that analyzes incoming emails and categorizes them based on predefined types set during the setup process.

Q5: What are the potential cost implications of using ChatGPT with Make?
A5: Using ChatGPT within Make may incur costs based on usage. However, using the API for automation can be very cost-effective, often only costing a few cents per operation.