LegalMind
An intelligent platform designed to enhance law students' critical thinking through AI-supported case analysis. The tool uses stealth assessment to measure and scaffold legal reasoning skills without disrupting the natural learning flow. Features a rubric-guided AI evaluation system and personalized feedback engine.
ProductiveMath
PI: Dr: Seyedahmad Rahimi
A human-in-the-loop generative AI web application enabling teachers to create algebra problems that support productive failure pedagogy. Compares small vs. large language models for problem quality, uses a custom rubric for automated assessment. Recognized with a Best Paper Nomination at AIED 2025, Palermo, Italy.
SmarterFlash
A generative AI-powered adaptive flashcard application studying self-regulated learning. Research explored how students balance AI automation with personal agency in knowledge review — revealing that learners spend significant cognitive energy preparing rather than passively consuming AI-generated content. Presented at ICLS 2026, Irvine, CA.
Metalingual
Co-designed an AI-powered platform for enhancing English language learning with a focus on TOEFL speaking preparation. Integrated AI-driven personalized instruction and feedback loops. Conducted user testing to refine functionality and improve user engagement. Presented at AECT 2024, Kansas City.
This project is a prototype. For more information, visit the project website.
Visit Project Website →
Clues in the News
Designed a game-based stealth assessment system that measures K-12 students' critical thinking skills through news analysis tasks — without students knowing they're being assessed. The system embeds assessment seamlessly into learning activities, eliminating test anxiety while gathering rich performance data.