From Chatbots to Coach Bots: Practical Uses of Generative AI in Higher Education
General description of the project
NOTE: This presentation had been helped by AI in various forms.
This presentation illustrates advances achieved in several of my courses through the pedagogical use of Generative Artificial Intelligence (genAI). Although no formal quantitative data are yet available to substantiate these innovations as best practices, qualitative indicators—such as students’ engagement, curiosity, and enjoyment—suggest that genAI integration has had a distinctly positive impact.
My experimentation with genAI began with automating the creation of exercises and examinations. Over time, this evolved into the development of complete web-based course materials generated through textual prompts and iterative guidance of AI systems. More recently, I have created an AI-powered “coach bot” designed to support students in developing academic projects. I am also exploring systems that assist programming students by generating adaptive strategies, offering tailored advice, and detecting excessive similarity between versions of submitted code.
In parallel, I have experimented with AI-enhanced conversational systems to enrich student dialogue and collaborative content creation. This work involves the evaluation and comparison of a variety of AI tools currently available to the academic community. Alongside the widely used chatbots—ChatGPT, Gemini, Claude, DeepSeek, Mistral, and Copilot—other tools have proven valuable for integrating multimodal materials, synthesizing textual content, and supporting both faculty and students in organizing, summarizing, and connecting diverse data sources.
In essence, this ongoing experiment aims to explore how genAI systems can be scaffolded to facilitate meaningful learning experiences and lighten students’ cognitive and logistical load. I will showcase practical examples of such integrations across disciplines and levels of higher education, demonstrating how AI can empower educators to move beyond the constraints of traditional LMS environments and design authentic, dynamic, and student-centered digital learning spaces.
Technologies
The technologies used encompass a range of generative AI platforms—notably ChatGPT, Gemini, Claude, DeepSeek, Mistral, and Copilot—along with specialized text integration and productivity tools for synthesizing and organizing instructional materials. These systems were implemented to automate routine academic tasks (exercise creation, website design, feedback generation), to assist students in project development through conversational interfaces, and to enable richer classroom discourse.
Effectiveness has been observed through increased student participation and engagement, improved material quality, and greater instructional agility, as instructors can rapidly prototype and refine resources.
Explain project results
The use of genAI has contributed to the institution’s teaching mission by fostering pedagogical innovation and demonstrating viable pathways for integrating emerging technologies into course design. For students, genAI has provided personalized academic support, facilitated more interactive learning, and helped cultivate critical AI literacy—a key competence in contemporary higher education. Faculty, in turn, have benefited from time savings and creative augmentation in instructional design and assessment.
Why it should be considered best practice?
This initiative exemplifies best practice because it:
Integrates AI tools strategically and ethically into diverse teaching contexts, beyond superficial experimentation;
Models faculty-led innovation grounded in reflection and iterative design;
Balances automation with human oversight, maintaining pedagogical intent while leveraging technological potential.
Its replicability lies in the modular and scalable approach—any instructor can adapt the framework to their own discipline, course format, and institutional resources.
Highlights of your proposed presentation
The presentation will highlight:
The progressive evolution from task-level automation to AI-driven course ecosystems;
The design and function of the coach bot and AI-assisted programming tutor;
Examples of student engagement and creative outcomes;
A taxonomy of AI tools for academic use.
Key lessons learned include the importance of transparency and communication with students, continuous evaluation of tool reliability, and pedagogical alignment between AI use and learning outcomes. Above all, the experience reaffirms that the real innovation lies not in the tools themselves, but in how educators reimagine teaching practices through them.
The Evaluation Committee will evaluate submitted proposals based on the following criteria. Each area will be rated on a scale from 1 to 5 (1= non-satisfactory; 5 =outstanding), for a maximum of 45 points.