As a digital education professional, I took an exciting step today by creating my first custom GPT during my lunch break. The process turned out to be surprisingly intuitive and took only about 10 minutes from concept to completion.

The Problem I Wanted to Solve

In my role, I handle numerous digital education queries daily, primarily through email conversations with users. While I maintain a GitHub repository to track these issues, creating comprehensive summaries has always been time-consuming and challenging. I needed a more efficient way to convert lengthy email threads into concise, well-structured issue summaries.

The Solution: Custom GPT

After experimenting with ChatGPT and seeing how effectively it could summarize email conversations, I realized creating a specialized GPT could streamline my workflow. This led to the development of “DigEd Query Summariser,” a custom GPT designed specifically for my needs.

Creating the GPT: A Simple Process

The development process was remarkably straightforward:

  1. I started by explaining my concept to the ChatGPT interface
  2. The system guided me through each step, suggesting names and helping me refine the functionality
  3. I could test and adjust the GPT’s responses in real-time
  4. The entire process felt like having a natural conversation with a helpful assistant

Future Improvements

The beauty of custom GPTs is their flexibility. I can continue to refine my tool by:

  1. Enhancing the Context

    • Fine-tuning the summarization instructions
    • Adding specific requirements for query details (e.g., requester, school, follow-up status)
    • Improving the accuracy and consistency of summaries

  2. Customising the Experience

    • Adjusting the name or profile picture to better reflect its purpose
    • Adding more specific prompt starters to guide users
    • Refining the response format based on actual usage

The Impact

This custom GPT has already started to transform my workflow. What used to be a manual, time-consuming process of summarising email threads for GitHub issues is now streamlined and consistent. It’s a perfect example of how AI tools can be personalised to solve specific workplace challenges.