- Constructed LGBM models with direct recursive hybrid strategy to forecast sales 9 months ahead, obtaining under 15% error during pandemic periods and enhancing average 30% allocation precision. - Built the ATP optimization tool with Linear CBC (Branch-and-Cut) Solver by the python ortools package on a scale of some 4,000 lines, speeding up weekly routines of 2-day workloads to within 1 minute. - Ran a PoC of CI/CD on cloud platforms inclusive of AWS, GCP, and OCI, contributing to comparison report evaluating 3 platforms through time, cost, EoU, IAM, compatibility, etc.

- Lectured in math with up to 40 students at a time, helping them get progress in their grades. - Wrote high school math textbooks about data analysis and probability with 4 times efficiency compared to other teachers, assisting the cram school in adapting to the education reform better.

- Applied K-means to extract more information about FunNow users, resulting in 3 clusters for the marketing team to establish proper promotion strategies. - Crawled other e-commerce competitors’ data like gomaji.com, myfave.com, and tripadvisor.com. by requests and selenium, boosting contracts of the business development team to reach weekly record levels: 20 per BD. - Automated organizing users' browsing data on Amplitude (BI platform), structuring 2 million data every month. - Used MySQL with Metabase to create useful dashboards for managers, supporting them to save 20% time spending. - Created a GUI of promotion code generator by tkinter, help the product team obtain 2 more months to develop a complete promotion code system on web.

- Applied non-linear GAM, KNN imputing, ARIMA, and XGB to 3 million data of Lending Club (the biggest P2P lending platform), obtaining pseudo r-squared 75% surpassing a paper in NCCU. - Crawled data in lnb.com.tw and lend.com.tw by requests and selenium with cracking CAPTCHA techniques, contributing to forming 100% data of the company. - Implemented a trading strategy with time series methods on financial data, boosting the company to increase 5%~10% profits in the year. - Cooperated with back-end engineer in building PostgreSQL DB on GCP, allowing financial data derived from TEJ and Reuters Eikon to be stored. - Setting up SOP for reading JCIC reports, helping coworkers systematically check users' credit information.