Transform the Trajectory of a Student’s Life, Support Outstanding Faculty Research, and Create Tomorrow’s Leaders



Transform the Trajectory of a Student’s Life: Support ARCS’ Internship Program



California State University, Northridge

Research Project

AI Mentor

Research Team

Lead Researchers:

  • Nhut Ho, Mechanical Engineering

Collaborators:

  • Dr. Thomas Lu, Rema International

Student Team:

  • Arthur Lazaryan, BS, Computer Science

  • Sara Madjdi-Sorkhabi, BS, Computer Science

  • Arrshan Saravanabavandam, BS, Computer Science

  • Devansh Sharma, BS, Computer Science

  • Heet Patel, BS, Computer Science

  • Vikas Sardhara, BS. Computer Science

  • Abdul Imran, BS, Computer Science

  • Timothy Do, BS, Computer Science

  • Jose Tobon Rodriguez, BS, Computer Science

Funding

  • Funding Organization:
  • Funding Program:

SYNOPSIS

The AI Mentor project is designed to support Mechanical Engineering students enrolled in ME 101 by providing an intelligent, course-specific learning platform. The system leverages artificial intelligence to deliver personalized study materials and interactive learning support tailored to each student’s academic progress and learning style. By offering customized study sets and guidance, the AI Mentor enhances comprehension of foundational engineering concepts, promotes efficient study habits, and fosters long-term academic success. Ultimately, this tool aims to bridge the gap between traditional instruction and individualized learning, empowering students to strengthen their understanding of core topics and improve their preparedness for future academic and professional opportunities.

Abstract

The AI Mentor system leverages a suite of advanced technologies including, AI agents, Retrieval-Augmented Generation (RAG), Large Language Models (LLMs), and the Model Context Protocol (MCP) to generate personalized study materials, deliver targeted problem-solving guidance, and provide interactive feedback. Through the use of prompt engineering and dynamic knowledge retrieval, the platform adapts its responses to align with each student’s learning pace, comprehension level, and academic progress. By integrating state-of-the-art AI methodologies with sound educational principles, the project seeks to enhance learning efficiency, deepen conceptual understanding, instill the Conceive Design Implement Operate (CDIO) engineering framework within students, and to ultimately serve as a scalable model for AI-driven instruction within STEM education.

Motivation/Research Problem

Traditional teaching methods have long served as the foundation of education; however, as technology advances, students are now expected to absorb and apply larger volumes of information in shorter time frames. This increasing demand exposes the limitations of traditional instruction in providing individualized learning support. In foundational courses such as ME 101, students often face challenges grasping complex mechanical engineering concepts without personalized guidance, which can hinder confidence and academic performance.

According to academic sources such as the The American Society of Engineering Education, 50–60% of engineering students switch majors or drop out before graduation, with the first year showing the highest attrition rates due to difficult coursework and limited instructional flexibility. These trends highlight the urgent need for adaptive learning tools that can provide personalized, efficient support to students.

The AI Mentor project addresses this challenge by developing an intelligent, adaptive learning platform designed specifically for Mechanical Engineering students in ME 101. By leveraging artificial intelligence to tailor study materials, offer targeted explanations, and provide responsive feedback, the system aims to enhance comprehension, improve study efficiency, and reduce early attrition among engineering students.

Research Objectives

The primary objective of this study is to examine how the integration of Artificial Intelligence, evidence-based learning methods, and gamification strategies can enhance student learning outcomes. Specifically, the research aims to determine how these technologies can foster more effective, engaging, and cognitively enriching learning experiences for Mechanical Engineering 101 students. The objective will be reached by answering the following questions to form hypotheses which will help the group moving ahead with the project:

  1. How can Artificial Intelligence and AI-centric platforms be designed to help students learn more efficiently and effectively?
  2. How can these tools and platforms be structured to encourage students to think more deeply and critically about the engineering topics they engage with?
  3. What learning methods and cognitive strategies can be incorporated into the platform to promote deeper critical thinking, problem-solving, and fundamental topic understanding?
  4. Can the gamification of the learning process serve as an effective psychological motivator to increase student engagement, and ultimately their education outcome?
  5. How can Large Language Models (LLMs) and modern Artificial Intelligence strategies be utilized to support Mechanical Engineering 101 students in achieving meaningful learning outcomes over the course of a semester?
Research Methods
Data for this study is being collected through questionnaires administered to current Mechanical Engineering 101 students. These surveys provide qualitative insights and firsthand perspectives from individuals actively enrolled in the course. Additionally, interviews have been conducted with former Mechanical Engineering 101 students now in their upper-division coursework, as well as with current instructors of the course, to obtain a broader range of viewpoints and professional reflections.

From a technical standpoint, the research also examines emerging industry standards in the development of large language model (LLM) based systems, including Retrieval-Augmented Generation (RAG) and integration techniques facilitated through the Model Context Protocol (MCP). We are also exploring advancements in Agentic AI approaches and employing iterative prompt engineering to refine and optimize student interactions on the platform. Furthermore, we have analyzed existing educational technology offerings to assess their strengths, shortcomings, and the challenges they present when integrated into the established academic ecosystem familiar to students.

Collectively, these efforts will provide a comprehensive understanding of student user needs, technological necessities, and current constraints, which would guide us in the development of a platform which best suits the needs for students in Mechanical Engineering 101.

Research Results and Deliverables
The project is currently in its developmental phase, and a tangible product has not yet been produced; therefore, no concrete results are available at this stage. The completed platform is projected for release in early Spring 2026 and will undergo testing within multiple sections of the Mechanical Engineering 101 course offered during that semester. This implementation will enable the collection of authentic user feedback from students actively engaging with the system, allowing for iterative refinement and improvement of the platform based on their experiences. Once user feedback has been analyzed and platform utilization is underway, the findings will serve as the foundation for a comprehensive research paper documenting the system’s effectiveness, with the specific techniques and methods utilized to reach the research objectives.
Commercialization Opportunities
  • Applications: AI centric platform with the technologies and features to help achieve the key values of the project.
  • Key Values: Enhance engineering education, allowing more efficient and effective learning for students.
  • Potential Customers: Higher education institutions or professors specifically.
Research Timeline

Start Date: 05/09/2025
End Date: TBD

Research Team

Lead Researchers:

  • Nhut Ho, Mechanical Engineering

Collaborators:

  • Dr. Thomas Lu, Rema International

Student Team:

  • Arthur Lazaryan, BS, Computer Science

  • Sara Madjdi-Sorkhabi, BS, Computer Science

  • Arrshan Saravanabavandam, BS, Computer Science

  • Devansh Sharma, BS, Computer Science

  • Heet Patel, BS, Computer Science

  • Vikas Sardhara, BS. Computer Science

  • Abdul Imran, BS, Computer Science

  • Timothy Do, BS, Computer Science

  • Jose Tobon Rodriguez, BS, Computer Science

Funding

  • Funding Organization:
  • Funding Program: