• Designed deep learning compilers for AI inference accelerators

• Instructed two sections of 6 hours per week to roughly 55 students about probability, hypothesis testing, estimation, regression, and single variable calculus, creating biweekly quizzes to test students’ understandings of course materials.

• Taught students how to use R to manipulate data frames by base R or the dplyr package, make plots with the ggplot2 package, and use advanced functions such as lm, sapply, etc.

• Avoided overlooking potential financial crimes by conducting analysis on mortality grading, term NPR, and deterministic reserve, resolving discrepancies in 3 insurance companies’ reports, and presenting the result to the entire department.

• Boosted work efficiency significantly by 50% and eased compliance checks through generating reports that consist of key statistics, summary data, evaluations, and visualizations based on 35 insurance companies’ VM-31 reports.

• Created a guidebook according to the Actuarial Standard of Practice that standardized the working procedures of all actuaries.

• Scraped and cleaned data of political candidates, municipalities, and election results covering 14 years for local offices in California from the website SmartVoter in R to investigate the determinants of female representation in politics.

• Separated more than 45,000 observations into different spreadsheets from 1997 to 2010 and created functions in R to use regular expressions to classify candidates based on education levels.