Physics & Mathematics Education · AI-Guided Reasoning
I build AI-supported learning systems that strengthen student reasoning.
My work is motivated by a simple concern: if AI gives students answers too quickly, it can contribute to cognitive atrophy. I want to build systems that prevent that by keeping learners engaged in reasoning, explanation, revision and reflection.
My background combines physics education, mathematics, teaching and software engineering. I use that foundation to design and study AI-guided tools that support conceptual understanding without replacing the thinking students need to do themselves.
Direction
Current Direction
Cognitive Effort
Studying how AI can help students stay mentally active instead of bypassing the struggle that builds durable understanding.
Reasoning Support
Building tutoring systems that ask guiding questions, require explanations and move learners step by step through difficult concepts.
Learning Evidence
Using pre/post-test evaluation, learner interaction logs and qualitative analysis to study how understanding develops over time.
Evidence
Research and Systems Background
My research and systems work focuses on AI-guided support for reasoning, conceptual understanding, classroom evaluation and interaction data that can help explain how students think.
AI-Guided Reasoning Support for Conceptual Understanding
A controlled pilot study of an AI tutor that guides electric circuit reasoning step by step instead of giving direct answers.
This work connects physics education, learner interaction logging, pre/post-test evaluation and the design of AI systems that strengthen students' own thinking.
Circuit Quest: A Structured Interaction Environment for Reasoning in Electric Circuits
A completed interactive system for circuit construction, prediction, simulation and reflection.
This work shows my interest in building systems, collecting useful interaction data and studying how people learn from feedback.
Engineering Design and Cognitive Engagement in STEM
A study of how design-based learning can improve conceptual understanding in STEM classrooms.
This work supports my interest in applied systems, reasoning, experimentation and useful technology for real environments.