May 2018 – Aug 2018 · 3 mos · Greater New York City Area

• Created an error-grouping API based on TF-IDF vectors and Agglomerative clustering algorithm, which categorized more than 40k quantitative trading error messages per day and decreased manual-checking time by 98%
• For each cluster, applied NLP and N-Gram model to error messages and generated human-readable English summary
• Developed an Outlook plugin that users can use Unix-like commands to upload email bodies and attachments to Jira as comments, create Jira issues, and add/remove watchers

May 2017 – Sep 2017 · 4 mos · Pittsburgh

• Developed a Python web crawler scraping information from https://community.babycenter.com, including all topics, threads and comments with their authors, post time and IDs
• Set up two databases using PostgreSQL to store collected data

May 2017 – Aug 2017 · 3 mos · Pittsburgh

• Project Name: Best Optimal Investment Analysis of Stock Market
• Studied special conditions in stock market, under which investors want to invest a fixed portion of their initial capital into stocks and put the rest into money market to optimize utility of expectation of return
• Built Lagrange Multiplier Model (programmed in MATLAB) to generalize the result to infinite stocks with different utility functions
• Research result selected to Dr. Steven E. Shreve’s Stochastic Calculus for Finance Volume II http://www.math.cmu.edu/~shreve/

Apr 2017 – Aug 2017 · 4 mos · Pittsburgh

• Winner of XPRIZE (Global Learning) Finalists
• Analyzed 70k Json data from facial recognition and build a MySQL database for management

Jan 2017 – Jan 2017 · 0 mo

• Winner of Best VR/AR Hack (PennApps Route) and Best Space Hack
• Developed an educational VR Android app for user to travel in the solar system with Unity and C#
• Used Watson Unity SDK & Windows Phrase Recognition System to achieve Voice control