Introducing Finalist – Team Cogniva, Bandung Institute of Technology

This Institut Teknologi Bandung project applies learning-by-teaching with agentic AI to strengthen students’ self-regulated learning. Instead of receiving explanations, students teach concepts to an AI agent, which prompts reflection, planning, monitoring, and evaluation of their own understanding. By casting the learner as teacher, the system encourages metacognition and ownership of the learning process, key components of self-regulated learning. The agentic AI adapts its questioning to expose gaps and reinforce mastery, helping students become more independent, reflective learners. The project demonstrates how learning-by-teaching, powered by AI, can build durable self-regulation and deeper conceptual understanding.