About Me

Academics:

B.S. Computer Science at UC-Davis (2019-2022):
* Completed 4-year Computer Science program in 2.67 years (1.33 years early).
* Worked as a software engineer at CodeLab, a UC-Davis student organization that builds software projects for startups. CodeLab is UC-Davis' first software consultancy club.

Newport High School (2015-2019):
During my time at Newport High School, I took 3 advanced college level math classes (Advanced Calculus (Calculus III and IV) at Newport and Linear Algebra (MATH 208) and Differential Equations (MATH 238) at Bellevue College), and scored in the top 1% of all test takers for the SAT.

Test Scores/Coursework: AP Computer Science A (5/5, 9th grade), AP Calculus AB (5/5, 10th grade), AP Calculus BC (5/5, 11th grade), SAT Math Level II (800/800, 11th grade), As in Linear Algebra and Differential Equations (12th grade)
Languages I speak: English, French, Mandarin, and Shanghai Dialect
Skills: C, C++, Java, Python (Django, XGBoost, Sckit-Learn, NLTK), JavaScript (NodeJS, SocketIO, Firebase, Mapbox GL), HTML/CSS, Heroku Deployment, Git Version Control, UNIX OS

Work + Research

Software Engineer, Messaging Systems at Meta

May 2022 to present: Working on performance optimization on Android devices for Msys, Meta's centralized and cross-platform messaging library.

April to May 2022: Underwent Meta's Engineering Bootcamp: Learned about Meta's strategic vision, internal tooling, developer workflow + infrastructure, and explored Meta's variety of teams.
Software Engineering Intern at Microsoft

June 2021-Sept 2021: Cloud + AI Division: In the Platform Engineering Team, I am advancing the industry's standard for testing ABAP code, by developing an ABAP interpreter + transcoder which can auto-generate ABAP code coverage reports and unit test code. This product left-shifts ABAP testing, allowing enterprise products to be shipped faster.

* Technologies Used: C#, ANTLR, Microsoft IntelliTest
Undergraduate Research Assistant at UC-Davis

Mar 2021-Sept 2021: Conducting research to find the political biases of the world’s most popular social media platforms’ recommendation algorithms (Twitter, Facebook, Reddit, and YouTube) under Professors Zubair Shafiq and Magdalene Wojcieszak. Developed a novel algorithm to classify political biases of articles and wrote Python scripts to parse through millions of Twitter users, taking advantage of high-performance computing techniques such as multithreading.

* Technologies Used: Python, High-Performance Computing (multithreading and concurrent programming)
Software Engineer at CodeLab

Oct 2020-Jan 2021: CodeLab is a UC Davis student organization that builds software projects for startups. I collaborated with 4 other engineers to build an internal UI CRUD tool for Ambii to easily and quickly query and update their database through React + Ambii's internal backend APIs.

* Technologies Used: React, Ambii's internal backend APIs.
Software Engineering Intern at Microsoft

Jun 2020-Sep 2020: In the Enterprise Security Team, I worked on developing an intelligent change monitoring system that auto-emails managers/application administrators of employee role assignment changes and asks for their acknowledgement. Role change data is neatly displayed onto a PowerBI report.

* Technologies Used: C#, Azure Cosmos DB, Microsoft Graph API
Undergraduate Research Assistant at UC-Davis

Feb 2020-May 2020: Pre-processing data and training a deep learning neural network to identify swine farm locations based on satellite imagery, given data sets of known swine farm latitude/longitude coordinates.

* Technologies Used: Python, Google Inception V3 (CNN + Transfer Learning)
Software Engineering Intern at Microsoft through Pactera

Jun 2019-Sept 2019: Programmed automation technology that tags developer tickets through Python (Scikit-Learn, XGBoost, NLTK) and SQL:
* Worked for Microsoft through a contract from Pactera Technologies
* Learned and applied RNN and LSTM architecture to increase ML model performance.
* Gave input on how to streamline end-to-end production pipeline.

* Technologies Used: Python (ML libraries: Scikit-Learn, XGBoost, NLTK) and SQL
Software Engineering Intern at Microsoft (High School)

Jun 2018-Aug 2018: Programmed two projects: obfuscating customer data for data mining w/ GDPR concerns and using ML (K-Means+Word2Vec) through Python (Sckit-Learn) to auto-classify database incidents. Feel free to check out the slides detailing my work.

* Technologies Used: Python (ML libraries: Scikt-Learn, Algorithms: Word2Vec + K-Means)
Research Assistant at UW-Tacoma

Jun 2017-Aug 2017: Co-Authored a CS paper with professors from UW-Tacoma titled, "Embedded Middleware for Distributed Raspberry Pi Device to Enable Big Data Applications", focused on distributed embedded system architecture and implementing IOT systems using the Rasberry Pi. It was published in the 15th IEEE/IFIP International Conference on Embedded and Ubiquitous Computing.

* Technologies Used: Rasberry Pi, Python

Honors / Awards

USA Computing Olympiad Gold Qualifier

2017: Achieved a perfect score (1000/1000, placing 2nd/579) in the USA Computing Olympiad Silver Division Open Contest.

Publications / Research

Research Paper Co-Author

2017: Co-Authored a CS paper with professors from UW-Tacoma titled, "Embedded Middleware for Distributed Raspberry Pi Device to Enable Big Data Applications", focused on distributed embedded system architecture and implementing IOT systems using the Rasberry Pi. It was published in the 15th IEEE/IFIP International Conference on Embedded and Ubiquitous Computing.

Projects

DiffFinder: File/Text Comparison Tool

November 2019-present: DiffFinder is an Android app that highlights the differences between two versions of a file (plain text, Docx, or PDF) or two versions of inputted text. It has the following features:
* Clean and intuitive user interface
* Lists the number of additions + removals made to the original file
* Adjustable colors for addition and removal highlights
* Adjustable font size

DiffFinder is an ideal utility for those who edit documents frequently and forget what document they have saved as v1, v2, etc. It is also ideal for developers and software engineers to check changes in their code.
Based on user feedback, implemented an improved version (DiffFinder 2). In total, both apps accumulated 6,200+ downloads from 176+ countries with high ratings to match.
View Project

Ambii Internal Dashboard

Oct 2020-Dec 2020: Built an internal UI CRUD tool for Ambii to easily and quickly query and update their database using React and calling backend APIs.
View Project

Chatiton.com: 3-D Map-Based Anonymous Chat

Aug 2019-Oct 2019: Chatiton.com is an anonymous map-based chatting platform that seamlessly integrates a 3-D UI through Mapbox GL JS. Written in Node.js and deployed through Heroku.
View Project

Mazebattles.com: Race to Solve Generated Mazes

Jul 2018-Sep 2018: Programmed a site using Node.js and Socket.io where users race against each other to solve mazes generated using's Prim's Minimum Spanning Tree Algorithm. Features a single player, one-on-one mode, and a maze solving feature that uses Breadth First Search.
View Project

Arshi Editor

Sep 2015-Nov 2015: Wrote a text editor using Python Tkinter package. Built a syntax highlighting algorithm using Pygments for Python from scratch. Features syntax highlighting, file switch/merge/heap, automatic formatting, different modes.
View Project