28 A thought partner and research assistant

Imogen Kelly, Postgraduate Student, Digital Technologies and Education, United Kingdom 

I used GenAI in the context of a postgraduate education module focused on AI in learning design and higher education. As part of this module, I was tasked with two major projects: creating a concept map to document my learning journey and designing an educational artefact. I chose to develop SimPatient, a custom-built conversational GenAI tool designed to simulate clinical interactions for students in nursing, midwifery, and physiotherapy. Throughout the project, I used NotebookLM to summarise relevant literature, critically reflect on simulation pedagogy, and then I refined the tool’s design. GenAI was used both as a thought partner and as a research assistant. I also relied on it for academic synthesis and writing support during the development of my reflective evaluation report, however the initial outline and points were my own.

The project represented an applied, critical, and creative use of GenAI in an educational design setting, allowing me to test how GenAI can augment learner autonomy, design thinking, and scholarly communication in a higher education context. As part of this project, I created my own Custom GPT using OpenAI’s ChatGPT Plus platform. This allowed me to design a tailored conversational agent – SimPatient – specifically for healthcare education contexts. Customising the GPT involved defining its purpose, tone, knowledge boundaries, and behaviour so that it could simulate realistic, emotionally responsive patient interactions for students in nursing, midwifery, and physiotherapy. Creating a Custom GPT gave me fine-grained control over the learning experience, letting me shape not only the content but the pedagogical rhythm of each conversation. It was a unique opportunity to blend learning design with AI prompt engineering, demonstrating how students can go beyond using AI tools passively, to actively create them for educational use. 

GenAI was integrated throughout my learning process as both a reflective collaborator and an academic co-designer. Initially, I used it to help scaffold my thinking around key educational frameworks like experiential learning, simulation-based pedagogy, and personalised AI learning. This was particularly useful when designing the learning logic behind SimPatient. I also used GenAI to interrogate educational readings, extract quotes from scholarly sources I uploaded, and compare them to my own developing ideas. When working on my concept map, I asked GenAI to confirm that key themes I took from readings were linked to my concept map.

As the project progressed, GenAI supported the analysis of my written reflective report by suggesting ways to improve clarity, integrate relevant theory, and strengthen the argument through academic citation. What made this process particularly valuable was how the role of GenAI shifted from being a research tool to an iterative design collaborator, offering feedback, posing questions, and surfacing implications I hadn’t previously considered. This deepened my learning far beyond surface-level use.

I asked GenAI to read my final assessment and the rubric to ensure I was on the right track. As part of this project, I developed my conversational AI chatbot, SimPatient, which I designed to simulate realistic clinical conversations. The idea was to provide healthcare students with a safe, feedback-rich environment to practise communication, clinical reasoning, and empathy. Built using OpenAI’s Custom GPT platform, SimPatient features discipline-specific patient personas, scaffolded difficulty levels, and structured reflection prompts based on Schön’s reflective practice and Kolb’s (1984) experiential learning cycle.

Inspired by simulation-based education in healthcare, it enables learners to engage in emotionally nuanced patient dialogues and receive structured, rubric-aligned feedback. SimPatient not only supports skill development but also promotes metacognitive awareness through guided self-assessment. Its design was informed by my collaborative work with GenAI, which helped me align pedagogical theory with practical tool development, which demonstrated how AI can be used not just for content delivery, but for experiential learning. 

By demystifying academic research, enhancing reflection, and reducing cognitive load, GenAI helped me feel more empowered and intellectually engaged

GenAI enriched my learning by helping me think more critically, structure my ideas, and refine both the SimPatient tool and my reflective writing. It supported deep learning through prompting reflection, expanding the scope of research, and helping me make connections across diverse educational theories. For example, when I referenced concepts like Vygotsky’s (1978) ZPD or Schön’s (1983) reflective practice, GenAI not only explained them clearly but helped me apply them meaningfully to my design process. It also provided structured feedback, allowing me to revise and strengthen my academic writing with greater confidence. Importantly, it modelled a kind of responsive, dialogue-based learning that actually mirrored the pedagogy I was designing into SimPatient. It acted as a meta-learning partner, not just helping me produce work, but helping me understand how I was learning. This kind of responsive, just-in-time support would be impossible to replicate in traditional academic settings. By demystifying academic research, enhancing reflection, and reducing cognitive load, GenAI helped me feel more empowered and intellectually engaged throughout the process. 
 

GenAI Tool(s) Used:

References

Kolb, D.A. 1984. Experiential Learning: Experience as the Source of Learning and Development. Englewood Cliffs, NJ: Prentice-Hall.

Schön, D.A. 1983. The Reflective Practitioner: How Professionals Think in Action. New York: Basic Books.

Vygotsky, L.S. 1978. Mind in Society: The Development of Higher Psychological Processes. Cambridge, MA: Harvard University Press.

Many thanks