Projects

Presented research findings at the UIUC Engineering Research Symposium (ISRS)

Research contributed to published work at CVPR 2025 on object manipulation

Synthesized hand-object interaction under the guidance of Professor Saurabh Gupta, who specializes in computer vision, robotics, and machine learning.

Built a forearm GoPro mount to collect fine-grain hand manipulation

Modified a transformer-based model for 3D hand reconstruction

Crafted and conducted experiments to test and analyze poses, using linear regression for distance mappings

Integrated accelerometer and gyroscope (IMU) data to capture degrees of freedom

Segmented and removed hand from images to train robotic policies

Lead group project to build a stock trading app for investors to hone their skills

Built mySQL database and created synthetic dataset with Python scripting

Used a Flask back-end to connect database and get real time stock data with Yahoo Finance API

Hosted website and database on Google Cloud Platform

Users could: create portfolios, buy and sell stocks, research companies, and compare investing performance

Developed a SwiftUI iOS app to catalog graffiti at UIUC

Saved picture, name, and description for each work with SwiftData

Used MapKit to display art locations

Swipe through all works in gallery view

Flask backend integrating FAISS and HuggingFace LLM for RAG-style responses

Web scraper + BeautifulSoup parsed 700+ pages from McKinley Health Center’s site

Embedded text using SentenceTransformers and queried via semantic similarity

Built Python program to generate jet fuel and commodity market report with MatPlotLib charts/graphs

Data from XML based SOAP API, JSON based REST API, and web scraping with Beautiful Soup

Shared results weekly using automated email report pipeline created with Microsoft SharePoint and Power Automate

Led multi-functional teams to design app workflow and sampling process

Built Microsoft Power App to manage refinery sample inventory and lab tests

Created robust input forms interface with a Dataverse database

Data then feeds into AVEVA PI Vison or can be downloaded into Excel

Built interactive Svelte front-end for AFLEET Online’s Electric Vehicle (EV) calculator

The back-end is called with the user's input parameters

Returned calculations are used in a table developed to match the Excel model’s layout

Built the page according to the US Web Design System standards

The SQL query of the USURDB (United States Utility Rate Database) for EV charging not returning invalid information

Modified query and wrote a Python script to identify missing rates

Analysis found the USURDB database itself contained incorrect information

Notified the USURDB and correct rate information was published

Developed algorithm to calculate EV charging cost based on varying electricy demand and charger current draw

Users selected their local electrical rate, EV battery capacity, and charger for to compute their charging cost estimate

A data validation drop down was developed to build a dynamic range of electrical rates based on the state and type input