Testing Iris in Practice: Can AI Support Reflection and Sustainable Performance in Mentoring?

Iris is our WhatsApp-based AI Research & Thinking Assistant, and it has received approval from the Ethical Committee of Eindhoven University of Technology (TU/e) for use in this research. What makes Iris different is that it combines research measurement with evidence-based feedback, making it valuable not only for data collection but also for mentors’ self-reflection.

Iris:

✔️ Uses validated psychological scales

✔️ Generates personalized AI-based reports

✔️ Supports reflection linked to sustainable performance

We have started testing the use of Iris in the MOZAIC mentoring program. In this study, we explore how mentors’ empowerment and stress influence mentoring outcomes—and whether those who actively engage with Iris have a stronger impact on mentees’ perceived program effectiveness and sustainable performance compared to those who do not.

Early signals are encouraging. Mentors report that Iris is easy to use and satisfying as a WhatsApp-based assistant, with an overall USE experience (usability, satisfaction, ease of use) averaging 4.3 out of 5.

For this pilot, this matters because Iris is designed to do two things at once:

✔️ Support evidence-based reflection (not just “chat”)

\ ✔️ Enable lightweight research data collection in a way that fits into daily practice

The next step is to examine whether engaging with Iris actually strengthens mentoring outcomes—and under what conditions, such as levels of stress and empowerment.

Institute Reflection: Why This Matters for AI in Learning and Development

At the Institute for Sustainable Human Performance, this pilot touches on a broader question we see across contexts: Can AI support learning without replacing the human processes that make learning meaningful?

Insights from our co-creation work on AI in learning and development suggest that the value of AI does not come from the technology itself, but from how it shapes reflection, attention, and behavior over time.

In this sense, Iris is not designed as an automation tool, but as a reflection trigger—a way to make thinking visible and structured.

From a research perspective, this connects with:

• The Job Demands–Resources (JD-R) model, where tools like Iris can act as resources that support self-regulation under pressure

• Self-Regulated Learning, emphasizing the role of reflection and feedback loops in learning

• Self-Determination Theory, highlighting the importance of autonomy and intentional engagement

What we are beginning to explore is whether minimal, well-designed AI interactions—embedded in real work contexts—can support meaningful development without requiring heavy data collection or intrusive monitoring.

References:

• Job Demands–Resources (JD-R) model: Bakker, A. B., & Demerouti, E. (2007). The Job Demands–Resources model: State of the art. Journal of Managerial Psychology, 22(3), 309–328.

• Self-Regulated Learning: Zimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. Theory Into Practice, 41(2), 64–70.

• Self-Determination Theory: Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268.