• Implemented subscription system and built web UI using Spring Framework for credit score update to help users monitor their financial health• Incorporated RabbitMQ as a message broker to achieve query buffer and decoupled the services• Developed Data-Updater to subscribe data from RabbitMQ to format data, trigger corresponding actions and persist data to MongoDB• Tracked and analyzed user behaviors by running A/B test on the subscription feature

• Applied exploratory data analysis and large-scale feature engineering on feature extraction and manipulation byusing LendingClub data in Python• Used Random Forest, Logistic Regression, and SVM algorithms in scikit-learn to help identify if the borrower will repay the loan by the mature date• Trained a multi-layer Neural Network model using TensorFlow to advance the prediction accuracy by 6.3 %• Identified and visualized top factors that influenced the prediction and created dashboards in Tableau

• Designed and developed Data Warehouse Dimensional & Relational Models, OLAP cubes, based on businessrequirements, to support high-end, mission critical online analysis, using Hadoop and AWS Cloud• Created ETL procedures to collect, parse, transform large-scale data sets, and store into Hadoop and other SQLdatabases• Diagnosed production issues, optimized warehouse for robust reporting, analysis and queries

• Analyzed stock data with time series models in order to rationalize investment strategies in turbulent financial market• Leveraged 10+ stocks in medical industry to develop feature analysis based on existing predictive models which improvedthe forecast accuracy by 2.1%