• Implemented backend API connections to fetch information from database with Prisma ORM• Designed and developed mobile frontend with NextJS and Expo Go• Applied Agile development practices and CI/CD pipeline, participating in sprint planning, code reviews, and cross-functional meetings to ensure project alignment and timely delivery• Conducted testing and debugging to maintain high code quality and ensure platform reliability

• Participated in a group of 6 and proposed the initiative to develop bias prediction service targeting current business operations• Applied Clean Architecture, Agile development practices and CI/CD pipeline for software development• Deployed ARIMA model in PyTorch for time-series analysis, predicting userbase growth with mitigated bias• Implemented filter for different backgrounds, observing a 70% increase on transactions with bias mitigated• Built backend in Flask for responsive user interaction and efficient processing

• Performed comprehensive desk research and independently authored detailed analyst reports for 4 tech sector stocks, encompassing financial statement analysis, market trends, competitive landscape and corporate business analysis• Conducted stock valuation on a company working in Intelligence Transport System, walking through the whole process of profit prediction, balance sheet control, stock price monitoring and P/E, EPS estimation• Actively communicated with mentors and peers on research methodologies, data compliance and working styles• Attended 60+ meetings with tech experts and buy-side asset managers and delivered meeting moments, deepening my understanding in the market landscape and product trends in the tech industry

• Extracted, transformed and loaded (ETL) financial data of S&P500 companies using database APIs• Conducted exploratory data analysis, categorical feature encoding and data oversampling and used tree-based feature importance analysis to select only 36 features among a total of 92 features• Developed and trained 5 machine learning models (LR, Decision Tree, KNN, Random Forest, XGBoost), tuned the parameters and filtered out our best model with ROC-AUC over 0.9 on training dataset and 0.75 over testing dataset, indicating high predictability and reliability• Secured 1st rank among 4 teams in the final presentation, showcasing leadership and technical excellence in a collaborative environment

• Implemented front- and back-end register/login functionality for GuessTheLocation project by establishing connection between server routers sending POST requests and database receiving POST requests• Acquired proficiency in both front- and back-end environments, including foundational skills in NodeJS and ReactJS• Collaborated with team members on implementing front-end register/login webpages using ReactJS and CSS

• Monitored blog speech every day and deleted speeches against Community Convention• Conducted internal oversight on administration actions (deletion/ban/warning) weekly through DBeaver Database• Upheld community independence and freedom of speech by suggesting modifications on Community Convention• Put forward 5+ proposals to improve user experience and users’ privacy and create amiable campus environment(The organization is later renamed as DanTa Team)

• Contributed to the comprehensive project lifecycle involving a coffee brand’s expansion abroad to the U.S. market, using Power BI and Tableau to visualize trends in pricing, logistics, and competitor strategies• Conducted thorough online desk research on primary competitors, suppliers and logistics and warehouse providers within the U.S. coffee supply chain, composing workpapers on 22 Excel spreadsheets with over 110 sheets• Proposed a business model by extracting insights using Python and SQL querying supply chain metrics, which later constituted the main structure of the final solution presented to the client• Collaborated on the final presentation, integrating insights and interactive Tableau visuals