Jump Simulation Center

Carle Illinois College of Medicine

AI Software Developer Intern
October 2023 - Present

Project: Spatial Computing for Medical Simulation Training in XR Devices

At Jump Simulation Center, I took part in an ambitious project that aimed to revolutionize medical training through the integration of spatial computing technologies with Extended Reality (XR) devices. The goal was to create highly immersive, interactive medical environments that replicate real-world scenarios, providing medical students and professionals with hands-on experience like never before.

To bring this vision to life, I developed the backend infrastructure using Node.js and Express.js, which enabled real-time interactions between trainees and the AI-powered chatbot. Powered by Natural Language Processing (NLP) and OpenAI's GPT-4 model, the chatbot played a crucial role by offering guidance throughout simulated medical procedures. It answered complex questions and provided key insights during critical moments. A big part of my role was fine-tuning the chatbot's responses by connecting it to a specialized medical database, ensuring accuracy and relevance during the simulations.

This project wasn't just about guiding users—it was about completely transforming the training experience. I worked closely with a team of healthcare professionals to design interactive 3D environments that were synchronized with the AI chatbot, allowing trainees to manipulate virtual medical tools and practice procedures in real-time. These XR-powered simulations offered a seamless learning experience, immersing trainees in realistic medical scenarios where they could experience the same pressure and precision as they would in real-life situations.

Another key aspect of the project was training a specialized Large Language Model (LLM) for medical purposes. This allowed the AI assistant to provide contextually relevant, adaptive responses during simulations, enhancing the overall depth and realism of the training. I worked closely with healthcare experts to continuously refine the system, ensuring that it met the stringent standards required for medical training.

From backend development to real-time AI integration, this project utilized a wide range of technical skills. I used Python to train the LLM, and Node.js and Express.js to build the chatbot infrastructure. Additionally, I employed XR technologies to develop the interactive simulations. This project not only pushed the boundaries of AI in medical education but also demonstrated how advanced technology can revolutionize training in complex fields like healthcare.

Technical Skills Used:

• Backend Development: Node.js, Express.js for AI chatbot infrastructure.
• AI Development: Python for LLM model training.
• Spatial Computing: XR development for interactive medical simulations.
• Database Management: Medical database integration for LLM training.
• NLP: Implementing real-time conversational AI for enhanced user interaction.

Social Computing System Lab

University of Illinois Urbana-Champaign

Research Assistant / LLM & XR
May 2024 - Present

Focus: Research Analysis on LLM Integration in XR Environments

As a Research Assistant at the Social Computing System Lab, I worked closely with PhD students to explore the potential of integrating Large Language Models (LLMs) into Extended Reality (XR) systems. My role involved conducting thorough literature reviews, analyzing cutting-edge research, and synthesizing findings to understand how LLMs could enhance the interactivity of virtual environments. We aimed to discover new ways to make XR experiences more responsive and adaptive to users, elevating the overall immersion and engagement.

Throughout this project, we investigated how LLMs could significantly improve the adaptability of XR platforms, making virtual environments more intelligent and interactive. The goal was to enable real-time, context-aware responses within these virtual spaces, transforming how users interact with virtual agents and the environment itself. By identifying gaps in existing technologies, we proposed solutions that leverage LLMs to offer richer and more dynamic user interactions.

As a co-author, I contributed to a research paper that we've submitted to the CHI 2025 conference. The paper showcases how integrating LLMs with immersive technologies could redefine human-computer interaction (HCI) by creating more engaging and intelligent virtual experiences. Our research demonstrates how LLMs can enhance not only conversational interactions but also decision-making processes within XR environments, resulting in more immersive and responsive experiences for users.

The research also highlights how LLMs can enhance not only conversational interactivity but also decision-making processes within the XR environment, enabling a more holistic and immersive user experience. Through rigorous analysis and collaboration, we identified key factors contributing to the success of this integration, laying the groundwork for future developments in the field.

Research Skills Employed:

• Literature Review: Comprehensive analysis of academic and industry sources related to LLM and XR technologies.
• Data Synthesis: Integrating diverse research findings to develop cohesive insights into the future of LLM in XR.
• Academic Writing: Drafting sections of the research paper that were submitted to CHI 2025.
• Collaborative Research: Working closely with PhD students to refine methodologies and conclusions, ensuring the rigor of the research.

View the Research Paper Submitted to CHI 2025

Social Computing System Lab

University of Illinois Urbana-Champaign

Research Assistant / Machine Learning
January 2024 - May 2024

Focus: Developing Custom Machine Learning Models for Real-Time Object Detection

As a Research Assistant at the Social Computing System Lab, my work centered around exploring different approaches to machine learning, with a particular focus on real-time object detection. Early on, I conducted a thorough literature review to understand the landscape of object detection techniques and identify a specific area of interest. After diving into the research, I chose to focus on YOLO (You Only Look Once) because of its impressive balance between speed and accuracy in detecting objects in fast-paced environments.

This experience wasn't just about building models; it was about learning how to approach research effectively. I spent a lot of time analyzing previous studies, understanding the strengths and weaknesses of existing models, and identifying areas where further exploration was needed. Collaborating with PhD students and researchers gave me valuable insights into how to critically evaluate the research field and develop a clear focus for my own work.

By choosing YOLO as my area of study, I was able to explore its practical applications in real-world scenarios like autonomous systems and robotics. My role involved working hands-on with Python and PyTorch to build, refine, and optimize YOLO models, applying what I had learned from my literature review. This allowed me to experiment with fine-tuning the model, adjusting parameters to improve its performance in dynamic environments.

Throughout this project, I gained valuable experience in preparing large datasets, training models, and testing their accuracy. More than that, I learned the importance of collaboration, frequently working with team members to align our efforts with the lab's broader research goals. My contributions helped further our understanding of how machine learning, particularly YOLO, could be integrated into larger systems that require real-time object detection.

This project was a turning point for me as a researcher. It taught me how to dig deeper into a specific field, identify gaps in the current research, and make meaningful contributions. The hands-on work with YOLO and machine learning models, combined with the collaborative research environment, gave me a solid foundation to build on in the future.

Research Skills Employed:

• Machine Learning Development: Built and optimized custom machine learning models for real-time object detection using YOLO.
• Data Preprocessing: Handled large-scale datasets, ensuring data was clean, well-annotated, and ready for model training.
• Collaboration: Worked closely with PhD students to integrate object detection models into broader system architectures, improving real-time performance.
• Model Optimization: Employed advanced optimization techniques using Python and PyTorch to improve the accuracy and efficiency of the detection models.

Illini Fuel

University of Illinois Urbana-Champaign

Nutrition Analyst Intern
August 2023 - October 2023

Focus: Optimizing Nutritional Strategies for NCAA Athletes

As a Nutrition Analyst Intern at Illini Fuel, I played a pivotal role in optimizing the nutritional strategies for NCAA athletes at the University of Illinois. My work centered around creating data-driven nutritional assessments to fine-tune both dietary intake and athletic performance. By analyzing performance metrics, I crafted personalized meal and hydration plans tailored to support each athlete's specific training and recovery needs.

Leveraging Python and Pandas, I conducted in-depth data analysis to identify patterns and trends in performance. This allowed me to make evidence-based adjustments to nutrition strategies, ensuring that each athlete's plan was aligned with their physiological demands. Whether it was optimizing macronutrient distribution or refining hydration approaches, my work helped athletes not only perform at their best during competition but also recover more effectively.

In addition to data analysis, I implemented an automated feedback system to capture real-time input from athletes. This system enabled me to make swift adjustments to their nutrition plans based on factors such as training intensity, fatigue, and recovery progress. By incorporating real-time feedback, I ensured that athletes' dietary strategies remained aligned with their ever-evolving training regimens.

Collaboration was a key part of my role as well. I worked closely with coaches and performance teams to ensure that nutritional strategies were fully integrated into the athletes' training schedules. This partnership allowed us to create cohesive plans that supported athletes during intense training periods and recovery phases alike. I also tracked each athlete's progress to ensure the effectiveness of the strategies and continually made adjustments for optimal results.

This internship gave me a deep understanding of how nutrition directly impacts performance and recovery. It also provided hands-on experience in developing dynamic, data-driven strategies to maximize the potential of elite athletes both on and off the field.

Technical Skills Used:

• Data Analysis: Used Python and Pandas for in-depth analysis of performance metrics to optimize nutrition strategies.
• Meal Planning: Developed personalized meal and hydration plans tailored to individual athlete needs and performance goals.
• Collaboration: Worked closely with coaching staff to align nutrition plans with training schedules and performance objectives.
• Automation: Automated feedback collection from athletes, allowing for real-time adjustments to nutrition strategies.

Watch a video about the role here

Social Computing System Lab

University of Illinois Urbana-Champaign

Research Assistant / Unity Developer
May 2024 - September 2024

Focus: Unity Development for HoloLens 2 in Augmented Reality (AR) Research

As a Research Assistant and Unity Developer at the Social Computing System Lab, I was deeply involved in creating immersive 3D environments for research using Augmented Reality (AR) on the HoloLens 2. These environments were designed to enhance collaboration and spatial awareness during studies with PhD students, with a focus on improving real-time interactions and making the experience more engaging for users.

My role primarily involved developing interactive elements in Unity using C#, where I worked to ensure everything responded smoothly and handled complex interactions without lag. Performance optimization was key, as these AR applications needed to function seamlessly in real-time research scenarios. I made sure that the systems were both technically efficient and capable of delivering an immersive experience during studies.

Collaboration was a central part of the process. I worked closely with PhD students and researchers to fine-tune the AR environments, testing them in real-world scenarios to gather feedback. This iterative process helped us ensure the applications were user-friendly, immersive, and added real value to the research being conducted. The project pushed the limits of what spatial computing can achieve in research, showing how AR can create highly interactive and impactful learning and collaboration environments.

By combining my technical skills with a deep understanding of spatial computing, I was able to contribute to a project that not only advanced AR technology but also demonstrated its potential for academic and real-world applications. The experience highlighted the power of AR to transform how we interact with virtual spaces in meaningful ways.

Key Responsibilities and Skills Employed:

• Unity Development: Built interactive 3D applications using Unity and C# for HoloLens 2, focusing on spatial computing and real-time interaction.
• Performance Optimization: Optimized AR applications to ensure smooth functionality during research studies, with a focus on minimizing latency.
• Collaboration: Worked closely with PhD students to gather feedback and refine the AR environments based on real-world testing and study requirements.
• Spatial Computing: Developed AR environments that enhanced collaboration and spatial awareness in research studies, leveraging the capabilities of HoloLens 2.

Illinois Data Science Club

University of Illinois Urbana-Champaign

Project Manager / Machine Learning
August 2023 - December 2023

Project: Google Stock Analysis Using Machine Learning

As a Project Manager for the Illinois Data Science Club, I spearheaded a comprehensive project aimed at analyzing and modeling Google stock data through the application of advanced machine learning techniques and data science tools. The primary objective was to explore historical stock data in depth, extracting actionable insights while developing models to forecast future stock price trends using cutting-edge data science methodologies.

The project began with extensive data wrangling, where I leveraged Pandas and NumPy to clean, organize, and preprocess the data, ensuring its readiness for complex analysis. We pulled a large dataset of Google stock prices from Yahoo Finance, spanning over two decades of financial history, which provided a rich foundation for our analysis.

To uncover trends and predict stock movements, we implemented a variety of machine learning models including linear regression, time-series forecasting, and moving averages. These models were designed to track long-term trends while filtering out short-term fluctuations. I led the team in the creation of rolling averages for closing prices, which allowed us to visualize these long-term trends more effectively. Additionally, we used data visualization libraries such as Matplotlib and Seaborn to present our findings visually, offering key insights through clear graphs and highlighting anomalies in the dataset.

A significant achievement of the project was the development of a predictive model, which we designed to forecast future stock prices based on historical patterns. By employing advanced techniques like moving averages and trend indicators, we gained deeper insights into Google stock's potential future behavior, allowing us to make informed predictions. These insights were invaluable in understanding stock market dynamics and helped refine our approach to financial data analysis.

Throughout the project, I ensured smooth collaboration among the team, aligning efforts to achieve high-quality outcomes. This experience deepened my expertise in machine learning, data analysis, and predictive modeling, while also honing my leadership and project management skills.

Key Responsibilities and Skills Employed:

• Project Management: Led a team of data scientists and machine learning engineers to ensure project milestones were met.
• Data Wrangling: Used Pandas and NumPy for data cleaning, preprocessing, and analysis of historical stock data.
• Machine Learning Models: Implemented predictive models using linear regression and time-series forecasting techniques.
• Data Visualization: Created dynamic visualizations using Matplotlib and Seaborn to uncover trends and outliers in stock prices.
• Collaboration: Worked closely with team members to combine technical skills and ensure seamless integration of machine learning models into the final analysis.

View project on GitHub

CS 124: Introduction to Computer Science I

University of Illinois Urbana-Champaign

Course Assistant (CA)
August 2024 - Present

As a Course Assistant (CA) for CS 124: Intro to Computer Science I, I play an essential role in helping students grasp the foundational concepts of computer science. Throughout the semester, I support students as they learn programming languages like Java and Kotlin, providing guidance on topics such as data structures, algorithms, and loops. Whether it's during labs or coding assignments, I'm there to offer hands-on assistance, break down complex ideas, and help them develop problem-solving skills. I also help students prepare for exams, ensuring they feel confident and prepared for success.

Beyond tutoring, I had the opportunity to contribute to a project that involved developing a Course Rating System for UIUC. This full-stack project allowed students to rate their courses and professors, providing valuable feedback to improve future course offerings. I worked on building the backend using Node.js, setting up the API to handle student feedback and store it in a MongoDB database. On the frontend, I collaborated on creating a user-friendly interface with React.js, making it easy for students to navigate and submit their ratings seamlessly.

In addition to my technical contributions, I collaborate closely with the course instructors and fellow TAs to continuously improve the learning experience for students. I'm committed to creating an engaging and supportive environment where students can thrive, helping them build a solid foundation in computer science that will carry them through their academic journey and beyond. It's been rewarding to see how these efforts, both in teaching and developing projects, have positively impacted the students and the course overall.

Key Responsibilities and Skills Employed:

• Tutoring and Support: Assisted students with understanding core concepts in Java and Kotlin, including algorithms, data structures, and coding best practices.
• Project Development: Collaborated on developing the Course Rating System using Node.js, MongoDB, and React.js to build a full-stack application for student feedback.
• Frontend and Backend Development: Helped implement both backend logic and frontend UI to ensure a smooth user experience for the course rating system.
• Collaboration: Worked closely with course instructors and other TAs to ensure students received timely assistance and support throughout the semester.

MOONBOOW Studio

Research Assistant (Software Developer)
November 2023 - January 2024

Project: Mixed Reality Experience on Meta Quest 3 with GPT-3.5 Turbo Integration

At MOONBOOW Studio, I was part of an exciting project that combined cutting-edge Mixed Reality (MR) technology with GPT-3.5 Turbo to create an immersive, interactive experience. Working closely with a team of developers and researchers, I helped design the backend infrastructure using Node.js and Express.js. My focus was on ensuring that the system could handle over 100 API requests per second, supporting real-time communication between users and the AI-driven virtual agents within the MR environment.

One of my key responsibilities was leading the integration of OpenAI's GPT-3.5 Turbo into the project, allowing the virtual agents to engage users in natural, seamless conversations. This conversational AI added a dynamic element to the MR experience, making interactions feel more intuitive and engaging. In addition to the backend work, I also contributed to developing interactive 3D elements in Unity, ensuring smooth transitions between the virtual and real worlds. These transitions were crucial in creating an experience that felt fluid and immersive for users.

To manage user interactions efficiently, I worked on a MongoDB database that handled over 10,000 records. Optimizing the database was key to ensuring fast performance and smooth access for all users. Throughout the project, I collaborated with UX researchers, conducting rounds of user testing to gather feedback and improve both functionality and user experience. This iterative process helped us refine the system, ensuring that the MR environment was not only technically sound but also engaging and user-friendly.

Together with my team, I co-authored a research paper that explores how integrating Large Language Models (LLMs) like GPT-3.5 can enhance user interactions in MR environments. The paper, which was submitted to the prestigious CSCW25 Conference, offers insights into how LLMs can make virtual agents more responsive and adaptable. Our research highlights the potential of LLMs to transform virtual experiences, making them more interactive, intelligent, and immersive for users.

Technical Skills Used:

• Backend Development: Node.js, Express.js for high-performance API management.
• Unity Development: Created and optimized 3D elements for MR experiences.
• Database Management: MongoDB for managing over 10,000 user interaction records.
• API Integration: Implemented OpenAI GPT-3.5 Turbo for real-time conversational AI.
• Performance Optimization: Scaled API and database performance for efficient real-time interaction.

Zhejiang Lvkang Medical Care Group Co., Ltd.

Software Developer Intern
June 2024 - August 2024

Project: Elderly Care Management Mini Program

During my internship at Zhejiang Lvkang Medical Care Group Co., Ltd., I had the opportunity to collaborate with a talented software engineering team to develop an Elderly Care Management Mini Program. This application was created to streamline the care management process in senior care facilities, giving caregivers an efficient way to track and manage the daily needs of elderly residents. The goal was to improve the quality of care by providing caregivers with a user-friendly tool to log critical information and respond to residents' needs in real-time.

In my role, I helped build key features of the mini program using JavaScript, the WeChat Mini Program framework, and Node.js. These features allowed caregivers to easily update and track essential information like medical appointments, dietary preferences, and activity schedules, all in real-time. By simplifying this process, we made it easier for caregivers to stay on top of residents' needs and provide timely support. I also worked on backend optimizations, ensuring the system could handle data efficiently and that the user interface remained responsive even during high traffic.

Working closely with other developers, I played a significant role in improving the system's performance. We focused on optimizing database interactions and streamlining backend processes to ensure the program could handle large amounts of data while maintaining a fast, smooth user experience. This was critical for ensuring that caregivers could access and update information quickly without delays.

Throughout the project, I was also involved in testing and gathering feedback from caregivers who used the system. By incorporating their input, we made the program more intuitive and tailored to their workflow, ensuring it was easy to use while meeting the demands of a busy care facility. The end result was a more efficient, scalable application that helped caregivers provide better, more timely care to residents.

Technical Skills Used:

• Full Stack Development: Developed features using JavaScript, Node.js, and WeChat Mini Program framework.
• Real-Time Updates: Integrated real-time data logging and updates to enhance care management.
• Team Collaboration: Worked with a team of software engineers to ensure a smooth and efficient user experience.
• Performance Optimization: Optimized both the frontend and backend for scalability and responsiveness.
• Testing and Feedback Collection: Conducted user testing and implemented feedback to improve the system's functionality and ease of use.