AmIWrite: Enabling Scalable One-on-One Handwriting-Based STEM Tutoring with an LLM-Powered AI Tutor

Abstract

This work presents an AI tutoring system based on handwriting interaction to support scalable one-on-one STEM learning. Powered by an LLM, the tutor interprets students’ handwritten work and problem-solving processes to provide targeted guidance and feedback. The project investigates how natural handwriting input can be integrated with generative AI to improve personalized and scalable educational support.

Publication
In CHI Conference on Human Factors in Computing Systems (CHI) 2026