Carnegie Mellon (with Professor Ken Koedinger)

Personalized Learning Squared (PLUS): Doubling Math Learning by Optimizing Tutoring from Training to Practice.

Plus is the first ever novel Dashboard-driven, in-class human-AI tutoring model with low-cost remote tutors in 3 schools.

My research priority is to develop Large Language Models (LLMs) to provide tutors feedback in training as well as during tutoring sessions.

1. AAAI AI4ED Conference 2024: Improving Assessment of Tutoring Practices using Retrieval-Augmented Generation DOI
2. Provide assessment and explanatory feedback during senario-based training lessons on short answer responses using LLM-facilitated Named Entity Recognition.Our team hosted the AIED Workshop
3. Best Demo Award in AIED2023, DOI and youtube video
4. The first scenario Demo with Fast API and bootstrap
5. A NER Labeling Feedback using react.js and express.js
6. NER Feedback: CEUR
7. Future Goals: I am currently seeking to create Language Model (LLM) models that offers feedback to tutors during tutoring sessions through analyzing tutor-student dialogues and transcripts. Additionally, I aim to provide a comprehensive tutor performance analysis to enhance their teaching skills.

AI-Powered Personalized Learning Experience (APPLE) (Canadian Government Funded)

APPLE is an ongoing applied research project conducted at Uforse Education Group Inc., with an overall goal to propose effective personalized teaching strategies that are AI-powered and data-driven.

Our platform shows uniqueness and innovation in its emphasis on personal growth, self-driven learning, and goal setting. Students’ academic achievement is our main focus, but at the same time we pay strong attention to cultivating a sense of purpose within students and providing holistic personalized support. Different from other educational technology, we will start with goal setting and then implement backward planning. APPLE attempts to provide learning recommendations that are goal-oriented, time- sensitive and holistic.

Carnegie Mellon University: Computational Models of Learning

This is the Learn Lab Summer School Research Project Mentored by: Danny Weitekamp, Napol Rachatasumrit, and Professor Chris MacLella

Explore the synergy of human and machine learning, using generative models to create practical simulated learning models. These models enhance theory and instructional design, drawing from psychology, cognitive science, and computer science to predict human behavior without relying on prior data.

Project Presentation: link

Student Automated Assessment System (Uplanner)

This research aims to streamline the process of selecting the right educational path for students by matching their profiles with course data using Machine Learning.

Recieved Research and Experiment Grant $$$

Scan QR Code with your phone to Try out the assessment:

University of Toronto DGP Lab (with Professor Ishtiaque Ahmed)

Designing Peer to Peer Online Training

Designed a Matching Algorithm using Neural Network for student and tutors.

Presented at University of Toronto Undergraduate Research Conference

ADHD Assessment (Comming soon)

working with an NGP Grant Application in process