• Integrated multiple data sources into a single data unified platform machine learning solutionstatus monitor and governance.• Developed Solution Catalog Python library which used for data scientist to batch query largeamount of ML solution and their dependencies.• Help data scientist deliver online models and variables which used to detect fraud transactionand explore potential business partnerships.• Collaborate with data scientists to select appropriate training data models evaluate cloudservice cost based on different data scales.

Customer Pickup Coordinator Application● Enhanced pickup user experience using Java, JPA, JSP, Java Servlet, ICEFaces, and Spring MVC which allowed users to assign and cancel pickups in a shorter time window and saved budget. ● Developed pickup notification service via Spring Boot, Spring JDBC, Redis, and JMS which notified the drivers for pickup on road status.● Deployed pickup microservices to Cloud and using Jenkins for CI/CD development. Automatic Pickup Service MiddleTire● Implemented automatic pickup and shipper query via SQL and JDBC to fetch millions of data from multiple data sources through iSeries DB2 and optimized query time.● Created unit tests for the middle tier service via JUnit, Mockcito, and Spring annotations so that unit test coverage has been improved. ● Modified scheduled pickup endpoint request and response type through JAXP and tested in SOAPUI.● Monitored and troubleshoot the production issue through Putty, App-Dynamic, and Splunk. APR Raw Failure Map Visualization Project● Developed Raw Failure Web Application via Angular 6 and TypeScript which used for summarizing raw failure data and visualizing raw failure stations on a map from multiple perspectives and dimensions.● Visualized raw failure density data on choropleth map via D3.js, WebGL, Three.js, Mapbox, Deck.gl● Built RESTful API for table summary and visualization by Java, Spring Boot, JPA, and H2 Database.

ImHealthy: Angular Web Portal Development – Clinical Management System• Developed ImHealthy web portal using Angular 5, Npm, Nx, ES6, and TypeScript which used by thousands of social workers and clinicians for clinical and health information management.• Improved system security and performance by adding authentication, cache and log analysis.Front-end:• Designed interactive web pages via Material, CSS, and RxJS which generated dynamic forms.• Implement health data visualization using D3.js, ng-charts and Google map to show aggregation.Back-end:• Created RESTful API for web portal via Node.js, Express, and TypeORM that handled HTTP request.• Built imHealthy data warehouse by combing PostgreSQL, MongoDB that provided quick speed for storing and querying patients’ questionnaire, social network, and demographic data.• Implemented the algorithm to calculate patients’ well-being score via TypeORM and Node.js.• Deployed imHealthy web service via Apache using AOT that improved performance.HealthyU: Angular Web and Mobile Service Development• Developed HealthyU website using Angular 5, Nx, TypeScript that provided health service.• Designed an interactive frontend by RxJS and Clarity which handled the asynchronized request.• Implemented dynamic questionnaire web app via ngSwitch, ngIf, ES6 and Promise object.• Created a cross-platform service through the web and mobile app via Ionic and Cordova CLI.

• Developed Wellness Pavilion System using Angular, Npm, Nx, ES6, and TypeScript which providedhealth information management service for eight departments in School of Health and Rehabilitation.• Designed the frontend via CSS, LESS, Ng-Prime, JavaScript and RxJS which visualized health data.• Created RESTful web service using Java and Spring Framework which provided high consistency.• Implemented patient management database via PostgreSQL, MongoDB that generatedthe query approach to SOAP note and computer adaptive testing approach to PROMIS questionnaire.• Analyzed the user log data via ElasticSearch, Kibana, and Logstash which helped decision making.• Built TeleHealth Web System with MySQL and PHP that implemented user survey and analysis.• Designed the frontend via jQuery, AJAX, Bootstrap and Google Charts which visualized survey data.• Built integration tests with Selenium and PHP WebDriver and guaranteed the fundamental functions.

• Developed SEMSearch Intelligent academic text mining system--Segmented and classified millions of Chinese text automatically with Java and Python using concept extraction--Extracted main points, key words and conclusions in text intelligently with LIBSVM and Weka and KNN combined with SVM machine learning models.• Provided continued maintenance and development for Grammar label marking system--Implemented cascading searching and relational searching function by Java, JSP and XML--The system has been applied to 100 people National Social Science Fund project team in Literature college