Research Intern
University of Texas at Austin
June 2024 - August 2024
Worked on developing a machine learning algorithm to classify methane super emitters from satellite data
University of Maryland | B.S. in Computer Engineering
Expected Graduation: May 2026
About Me
I am a junior at the University of Maryland studying Computer Engineering. My passion for technology began
when I built my first computer at 13, sparking my journey to understand how computers work from the ground up.
From electron interactions in doped semiconductors to the intricacies of computer language architectures, I am
eager to explore it all.
Last summer, I held a research position at the University of Texas at Austin, where I developed an image-labeling
tool to aid in identifying methane super-emitters. I had the opportunity to present this work at the American
Geophysical Union (AGU) 2024 conference this past December.
June 2024 - August 2024
Worked on developing a machine learning algorithm to classify methane super emitters from satellite data
August 2022 - Present
Developed the MPECWatch website using Sqlite3 and Bootstrap
September 2021 - May 2022
Developed software using OpenCV to enable cheap cameras to detect microplastics ∙ Developed a software model that would simulate optimal spacecraft landing areas using OpenCV
Collaborated with a team to assess challenges in Navy data access and security, focusing on role-based authentication and interdepartmental communication. Due to limited access to direct operational data and personnel, we relied on secondary research and insights from a single point of contact to develop scalable solutions. Our work emphasized budget forecasting, scalability assessments, and practical implementation strategies to align with the Navy’s operational priorities.
MPEC Watch is a web application that provides astronomers with valuable insights into minor planet data. Using Python, SQL, and Bootstrap, I developed a tool that queries a large SQL database and generates informative visualizations. The project focuses on individual station observatory browsers, where I'm responsible for creating user-friendly interfaces using Bootstrap. I've overcome challenges related to data cleaning and visualization, implementing efficient SQL queries and leveraging Matplotlib for data plotting. In the future, I plan to expand MPEC Watch to include additional features such as real-time data updates and more advanced analysis tools.
During my research at the University of Texas at Austin, I developed an efficient image labeling tool to streamline
the identification of methane super-emitters from satellite imagery. This tool leverages the Google Maps API to
query nearby locations based on satellite image coordinates (longitude/latitude), enabling intuitive labeling of
unmarked emitters. A key feature is its ability to batch images by geoproximity, allowing users to label clusters of
emitters in nearby areas.
In addition to increasing efficiency, the tool supports scalability for larger datasets, making it adaptable for broader
applications. I also worked with machine learning researchers to use the labeled
images for training a convolutional neural network (CNN), which achieved an accuracy of approximately 83% in identifying
methane super-emitter hotspots. The scientific community recognized my work, and I presented my findings at the American
Geophysical Union (AGU) 2024 conference in Washington, D.C. on December 7th.
As the leader of a five-member hackathon team, I spearheaded the development of an Arduino-based heartbeat simulator. This interactive educational tool uses LEDs within a 3D-printed heart model to visually represent a beating heart. A remote control allows users to adjust the heart rate via an IR receiver, providing a tangible and engaging experience while providing descriptors of the heart rate based on heigh, weight, and age. The project required extensive collaboration and communication among team members, as well as proficiency in Arduino programming and circuit design. I'm proud of the final product and the positive feedback we received from our peers.
For this project, I repurposed an old PC I built from spare parts during an upgrade. To make the system fit,
I had to get creative by installing mid-sized components into a micro ATX case. This involved purchasing a custom
cable with suitable wattage for the power supply, as it wouldn't fit in the standard location.
The server runs on a Ryzen 5 2600 processor, 16GB DDR4 RAM, a 1TB M.2 SSD (purchased specifically for the
server), a Corsair 550W PSU, Aorus B450 Elite motherboard, and an RX6600XT graphics card,
which I picked up used for about $200!
This self-hosted Nextcloud server manages my files, calendars, and contacts with complete control and privacy,
providing an efficient solution for my personal cloud needs. Running it on my custom-built PC allows for a highly
tailored and energy-efficient setup, while the unique design of the build highlights my hands-on problem-solving
and attention to detail in system configuration and hardware integration.