I am an undergraduate student pursuing a Bachelor of Science degree in Computer Science with a Mathematics minor at the University of California, San Diego.
I love learning new things and efficiently solving algorithmic problems. Throughout my past experiences, I've been able to explore many different fields within computer science, such as infrastructure software engineering, machine learning, and full-stack development. My favorite thing about this amazing field is the endless ways to think, grow, and make an impact.
One of my current endeavors is doing research in the field of supervised machine learning and computer vision, where I work with technologies such as OpenCV and PyTorch.
I have experience developing full-stack web apps using technologies and languages such as React, HTML, CSS, Node, MongoDB, Express, Firebase, Postman, and Jest.
I enjoy writing scripts and producing tools that help shorten software development life cycles. My experiences in this field include scripting in Python and Bash.
I am enhancing the experience of external and internal Google Cloud customers and users by analyzing aspects such as user traffic, errors, and latency, and designing, implementing, and improving load balancing pipelines and algorithms for these analyses in C++. I also make make improvements to the metrics used to identify load balancing improvements by designing and implementing crucial charts and queries.
I built an infrastructure tool that dropped the number of crashes/bugs on Meta products/sites and increased engineer productivity by automating code syncing between two of Meta’s largest mono repositories as they share many libraries and code. The tool shows code sync previews while code is under review to ensure the syncs will not cause crashes in the other repository. Once the original code is approved and pushed to production, the synced code is automatically pushed to production by this tool.
As an intern on Google Geo's Local AI team, I built a machine learning pipeline in C++ that efficiently filtered through billions of edits using MapReduce to generate positive labeled training data to retrain and potentially improve the accuracy of multiple machine learning models within Google Maps. Positive labeled training data is difficult to mine and previously only made up less than 8% of Google Geo Local AI team’s training data, my pipeline almost quadrupled the number of positive labeled training data.
In order to improve the WiFi chip and firmware quality from chip vendors, I designed and developed the infrastructure of a qualification tool for vendors to self certify their deliverables and serve as an internal sanity check tool that runs regularly to find any potential WiFi/Bluetooth connectivity issues for Google Nest’s Platform Connectivity Team. I also designed and implemented a graphical user interface to help vendors visualize the statistics collected from these tests, which included automatic emails with test summaries and Google Drive upload of detailed test logs.
I use medical imaging software, Pydicom, and the DICOM dictionary data structure to write Python software that increases the efficiency of sharing, viewing, and analyzing multi-layered flight and ground test data of U.S. nuclear weapons. I also created a Flask web-app to efficiently convert the data in large csv files to xkcd graphs.
I tracked baboon movements and behaviors through drone footage by applying concepts in machine learning and computer vision such as neural networks, Kalman filters, blob detection, mask generation, and centroid calculations. My work accelerated the Uaso Ngiro Baboon Project by simplifying the data to ease the analysis of baboon troop dynamics. In addition, I helped develop a new Agile inspired business model that makes the workflow at Engineers for Exploration more efficient. Our team was invited to the Measuring Behavior 2020 International Conference in Poland after publishing a paper on our research. Feel free to check out our paper on page 33 below!
I was a tutor for CSE 15L: Software Tools and Techniques (vim, linux, bash and shell scripting, unit testing, and version control) for 700+ students. I was also a tutor for CSE 100: Advanced Data Structures (C++). Some of my responsibilities included holding tutoring hours for one-on-one help, moderating labs in course subjects, grading assignments, and developing a bash script that is assigned to all students as a project.
I helped develop and moderate the official website and required tools for CSE 110: Software Engineering at UC San Diego. Tools on the website include forming teams based on student schedules, project submissions, team feedback and discussion, managing and moderating project schedules, and other needs in the UCSD software engineering course. Technologies used include React, Node, Express, sqlize, and GraphQL.
I was a mentor and project advisor in UCSD Summer Program for Incoming Students (in Computing). I helped teach Python and algorithmic problem solving to over 50 UCSD computing students, hosted lab hours for one-on-one help, and graded homework and labs. I also hosted networking events and panels that greatly benefitted the students. In addition, I advised multiple project groups in developing WebApp and machine learning projects and directly mentored four students in the many aspects of this five-week program.
I worked on a variety of software tasks involving the front end, back end, and software for the Fairygon website, app, and product. An example of one of my tasks was to help implement RESTful API for Siri Shortcuts to make requests to the server for the product to accomplish. This job also provided me with experience in the tech startup industry.
I created a MERN Stack application that acts as a social networking platform for developers. Upon logging in, users can view the profiles of other developers, create posts that appear in the feeds of other developers, and like or comment on the posts that appear in the user’s feed. Some of the technologies I used included MongoDB, Express, React, Node, Heroku, Git, Postman, and Bcrypt.
I was part of a team of four that developed Grosseries, which is an immersive VR experience in which users can immediately see how their grocery decisions impact the world. Grosseries' mission is to educate everyday shoppers about how their purchase choices can impact the environment around them. We aim to improve outcomes in sustainability awareness in a public desensitized from statistics, news, and even infographics. In the VR world, users are put in an outdoor grocery store on an island and adding grocery items to their shopping cart could potentially affect the beautiful nature around them. For example, purchasing a plastic water bottle causes plastic bottles and trash to wash up on the beach and the sky to turn orange due to the carbon emissions. This project was submitted to SDHacks and won Best Sustainability Hack, Best Hardware Hack, the echoAR Challenge, and the Wolfram Alpha Prize.
I was part of a team of three that created CaliAlert, which is a website created with HTML, CSS, and JavaScript where users input their phone number and zip code which we stored in Google’s Firebase. When the USGS Earthquake Catalog API sends a warning about a predicted or current natural disaster, we used the Twilio API to send a text email with safety tips and information about the natural disaster to people living in the zip code areas that are or could be affected by that disaster.
I am part of a team of four that works on EventNow, which is a website linked to an app in which organizations can input events that are then displayed on a map. Users can then look for events that they are interested in through keyword searches or simply clicking on nearby icons on the map to view the details of the event. We use HTML and CSS for the front end of the website, JavaScript (node), JSON files, and Google Maps API for the back end of the website, and Dart to develop the app.
I created an iOS and Android app using the programming language Dart and the UI framework Flutter that helps users track and analyze their daily water intake based on the user’s age and weight. This includes many sources of water intake other than water itself such as fruits and vegetables.