Evidence Driven Learning Innovation

How Custom GPTs Can Support Research Teams in Advancing Their Work: Meet EDLI Explorer

Written by:

Min Zhuang

Custom GPTs are advanced AI models tailored to meet users’ unique needs by providing targeted insights. To support researchers in their work, Caitlin Kirby and Stephen Thomas developed EDLI Explorer, a custom GPT designed specifically for the EDLI team.

EDLI Explorer serves as a knowledge hub, offering insights, summaries, and connections between projects related to teaching and learning. It makes EDLI’s resources more accessible and understandable for both new and experienced team members.

My Experience with EDLI Explorer

When I joined EDLI in early 2023, I found it challenging to fully grasp the team’s scope and ongoing projects—especially since I started in the middle of the academic year, when many projects were already underway. While my colleagues were supportive, it was a challenge for me to get familiar with the new environment and, due to my introverted nature, I hesitated to ask too many questions. As a result, it took significant time and effort to fully understand EDLI’s work, its strategies, and how various projects connected to the team’s broader goals.

When I first tried EDLI Explorer at the end of 2024, I realized how transformative this custom GPT could be in helping researchers understand EDLI’s work and integrate more effectively. Reflecting on my initial experience, I believe that having a tool like EDLI Explorer would have made a difference in helping researchers advance their work. Here’s how it helps:

  • Streamlining Onboarding – For newcomers, EDLI Explorer offers a centralized and interactive way to learn about the team’s work, scope, and goals, significantly reducing the time required to get up to speed. It acts as a one-stop resource for understanding the team’s mission and ongoing projects. For experienced team members, it serves as a valuable tool to refresh their knowledge of past initiatives and identify new opportunities for collaboration.
  • Clarifying Programs and Goals – EDLI Explorer breaks down complex projects into digestible summaries, helping team members quickly understand their purpose and scope. For instance, I asked EDLI Explorer to introduce the goals, history, and scope of the Colleges’ Online Learning Academy (COLA) program. The tool provided a clear summary and even included probing questions at the end of its response, encouraging me to explore the topic further.
Screenshot of a ChatGPT conversation with the custom GPT, EDLI Explorer. The user asks, “Tell me more about the goals, history, and scope of The Colleges’ Online Learning Academy (COLA) program.” ChatGPT responds with a structured explanation, covering: Goals: Supporting faculty in online and hybrid teaching through strengthening faculty capacity, promoting evidence-based practices, fostering a collaborative learning community, and enhancing student learning outcomes. History: Describes COLA's launch in response to the demand for high-quality online education during COVID-19, with development input from instructional designers and researchers, and notes its ongoing expansion and refinement. Scope: States that COLA provides professional development opportunities such as workshops, training modules, and one-on-one support for faculty.
A screenshot of a conversation with the custom GPT, EDLI Explorer
  • Identifying Engagement Opportunities – By analyzing project details, the tool highlights potential areas for collaboration, making it easier to align personal skills and research interests with team needs. EDLI researchers can even upload their CVs or research statements and goals, allowing EDLI Explorer to suggest relevant projects to explore. This process empowers EDLI researchers to take initiative on their own, fostering a more comfortable and self-directed exploration experience.

Limitations of EDLI Explorer

While EDLI Explorer is a powerful tool, it has some limitations:

  1. Lack of Real-Time Updates – Since EDLI Explorer does not update automatically, it may not always reflect the latest project information. Even if new files are added, the tool does not indicate whether the information comes from the most recent version or an older document.
  2. Accuracy and Detail Limitations – While EDLI Explorer provides general insights into EDLI’s projects, it does not always include specific details. For instance, when I asked for a list of ongoing projects and their respective leads, it could not provide specific names as it stated,  “EDLI Explorer does not have direct access to the specific project”. Additionally, there are occasional inaccuracies in its responses. For example, in the same conversation about COLA, EDLI Explorer incorrectly stated that the program supports faculty across a network of colleges, when in fact, COLA is designed to support graduate students. 
  3. Effectiveness Depends on User Input – Like other generative AI tools, the quality of EDLI Explorer’s responses depends heavily on the quality of the prompts. Users need to have a basic understanding of the subject matter to ask targeted questions and receive meaningful answers.

EDLI Explorer demonstrates the power of custom GPTs in supporting research teams. It promotes independent exploration in a low-pressure environment, allowing users to engage with information at their own pace. Its interactive and personalized features make it easier to navigate the complexity of a multidisciplinary organization like EDLI. Although it has its limitations, it still helps researchers integrate more quickly. In the future, team leaders can consider custom GPTs as valuable tools for providing adaptive support that enhances both individual learning and overall team efficiency.

Min Zhuang

,