Course project for CS543: Computer Vision. Collaborated with two other graduate students to conduct a comparative analysis of multi-object tracking with classical computer vision techniques and modern deep learning pipelines.
[Paper] [Code] [Video]Course project for CS598: Computing and Wellbeing. Collaborated with three other graduate students to investigate how text sentiment patterns in sector-specific online Reddit communities (subreddits) reflect and respond to national monthly unemployment rate fluctuations.
[Paper] [Code] [Poster]Course project for CS598: Systems for GenAI. Collaborated with four other graduate students to develop a containerized LLM model sharding system that simulates inference across virtual edge devices, enabling practical latency and efficiency tests on a single device with configurable constraints.
[Paper] [Code]Course project for CS415: Game Development (4 Credit). Developed a thematic 2D side-scrolling platformer game using Godot and publicly available assets. Implemented custom enemies and decoration for each season.
[Video]Maintained the UIUC Agroecosystem Sustainability Center (ASC) website through PHP template modifications and custom WordPress blocks. Changes support categorized news posts, MailChimp integration for newsletters, and ease of future updates through advanced custom fields and plugins.
[Website]Course project for CS411: Database Systems. Collaborated with three other students to develop a full-stack web application allowing users to view general events at UIUC overlaid on an interactive map. Utilized Python to scrape calendar data and designed a MySQL DB to store events, users, and locations.
Collaborated with two other engineers in State Farm’s annual internal Hackathon to create a predictive machine learning model in PyTorch combining tabular and image data to predict housing price. Achieved a 65% improvement in RMSE relative to price prediction over tabular-only or image-only models trained on the same dataset.
Course project for CS225: Data Structures & Algorithms. Collaborated with four other students to implement a graph algorithm (Dijkstra's, Floyd-Warshall) approach to finding optimal speedrunning routes for video game Elden Ring. Representing bosses as nodes and time as weight with logic for pre-requisite nodes, predicted times within 5% of real world records. Visualized with Python.
[Code]Course project for IS308: Race, Gender, and Information Technology. Developed an interactive website to visualize both granular (semester-specific) and cumulative (time-series) university enrollment data by race. Used Python to process 50+ CSV files containing 18 years of enrollment data sourced from the UIUC Division of Management Information (DMI).
[Website] [Code]
