• Develop a responsive business landing website using Taiwind CSS and React framework.
• Introduced Agile methods with GIT , Jira and Confluence to the company, resulting in significantly improved collaboration efficiency across the team and notable enhancements in productivity.
• Facilitate weekly online meetings to present web development demos, leading discussions on collaboration with designers and refining development workflows.

Crafted and executed the “Limited Time Promotion & Buy Now” business request by Java Spring, leveraging Redis and RocketMQ, elevating performance by 10 times from hundreds QPS to 10 thousands QPS range.
• Segmented order service into promotion and order microservices according to DDD(Domain Driven Design) principles; Implemented cache preheating for active promotions; Completed lock and revert inventory service via Redis and Lua script.
• Optimized optimistic-lock for inventory updates in MySQL database, solving the potential over sell problem while increasing data throughput by 80% from 416.9/sec to 755.4/sec, validated by JMeter concurrency testing.
• Strengthened application robustness by developing comprehensive unit tests, using Junit and Mockito, augmenting test coverage from 60% to 85%, as indicated by Jacoco.
• Collaborated in setting up and maintaining a CI/CD pipeline using Jenkins and Docker, integrated Kubernetes with AWS EKS, streamlining the development process and ensuring reliable deployments.

• Accomplished a Kaggle competition-styled project to predict wine prices by given wine features and review feedback, using Python Pandas and Numpy for data wrangling, NLTK for NLP.
• Applied PCA Analysis to reduce dimensionality from 45 to 28 components; performed Linear Regression (regularized by Ridge and Lasso) and Logistic Regression, evaluating the result by plotting ROC, AUC.
• Constructed the neural network model using PyTorch; Decision Tree and Random Forests using Sckit-learn, tuning and setting the tree-depth to 30 to gain the highest accuracy score.

• Designed and built a full stack MVC web app to recommend Spotify songs, with React for frontend, NodeJS and ExpressJS for backend Restful API, allowing user to search, filter and collect songs, hosted on AWS EC2.
• Engineered a region-based song selecter by allowing user to click on countries in an interactive world map to get songs, using React-simple-maps and React Hooks.
• Collaborated in user recommendation function to perform unsupervised machine learning technique to classify all songs based on K-means technique.