[Paper Review] Intelligent Tutor: Leveraging ChatGPT and Microsoft Copilot Studio to Deliver a Generative AI Student Support and Feedback System within Teams

Wei‐Yu Chen|arXiv (Cornell University)|May 15, 2024

Artificial Intelligence in Healthcare and EducationMedicine3 citations

TL;DR

This paper presents an intelligent tutoring system integrated within Microsoft Teams, leveraging OpenAI's GPT-4 via the ChatGPT API and Microsoft Copilot Studio to deliver real-time, personalized student support and feedback. The system dynamically adapts content based on learner progress and input, demonstrating potential to enhance student engagement and provide educators with actionable insights into learning patterns.

ABSTRACT

This study explores the integration of the ChatGPT API with GPT-4 model and Microsoft Copilot Studio on the Microsoft Teams platform to develop an intelligent tutoring system. Designed to provide instant support to students, the system dynamically adjusts educational content in response to the learners' progress and feedback. Utilizing advancements in natural language processing and machine learning, it interprets student inquiries, offers tailored feedback, and facilitates the educational journey. Initial implementation highlights the system's potential in boosting students' motivation and engagement, while equipping educators with critical insights into the learning process, thus promoting tailored educational experiences and enhancing instructional effectiveness.

Motivation & Objective

• To develop a scalable, real-time student support system within the Microsoft Teams environment using generative AI.
• To personalize educational feedback and content adaptation based on individual student interactions and performance.
• To integrate large language models with enterprise collaboration platforms to improve learning outcomes.
• To provide educators with actionable insights into student learning behaviors through AI-driven analytics.
• To evaluate the feasibility and impact of deploying a generative AI tutoring system in an institutional education setting.

Proposed method

• The system integrates OpenAI's GPT-4 API with Microsoft Copilot Studio to build a conversational AI agent within Microsoft Teams.
• Natural language processing techniques are used to interpret student queries and generate contextually relevant responses.
• The system dynamically adjusts educational content and feedback based on real-time student input and progress tracking.
• Microsoft Teams serves as the collaboration and delivery platform, enabling seamless integration into existing educational workflows.
• The architecture supports stateless, scalable interactions using RESTful API calls to the LLM backend.
• Feedback loops are implemented to refine response quality and adapt to learner performance trends.

Experimental results

Research questions

• RQ1 How can a generative AI tutoring system be effectively integrated into an enterprise collaboration platform like Microsoft Teams?
• RQ2 To what extent can AI-generated feedback improve student engagement and motivation in academic settings?
• RQ3 How does dynamic content adaptation based on learner progress affect learning outcomes?
• RQ4 What insights can educators gain from AI-driven analytics on student interactions?
• RQ5 What are the technical and operational challenges in deploying a scalable, real-time AI tutoring system in educational institutions?

Key findings

• The system successfully delivered real-time, personalized feedback to students through a conversational interface within Microsoft Teams.
• Initial implementation showed increased student engagement, with learners reporting higher motivation due to immediate response availability.
• Educators received structured insights into student learning patterns, enabling more informed instructional decisions.
• The integration of GPT-4 via the ChatGPT API with Microsoft Copilot Studio enabled scalable, low-latency interactions.
• The system demonstrated feasibility in adapting content dynamically based on student input and progress.
• The deployment model proved extensible and compatible with existing institutional IT infrastructures and collaboration workflows.