Google One

1. Implement PC application by Electron, ReactJS and Protocol Buffers.

2. Implement and deploy servers and RESTful APIs by NodeJS, MySQL, AWS dynamodb and AWS elastic beanstalk

3. Make contribution to game engine side by C#.

1. Implemented new features and fixed bugs for OncoKB, which is an oncology knowledge base with 2000+ visits per day.

2. Developed components for React front-end with TypeScript, MobX and webpack.

3. Evaluated google reCAPTCHA for client protection to block unauthorized actions like page scraping.

4. Implemented automatically visual regression test with Jest, puppeteer, Docker, and added CI test with GitHub Actions.

5. Refactored an internal platform and developed new features for it with React (with Hooks), MobX and Firebase.

6. Developed APIs and services for Spring Boot (with MySql) back-end by using Java 11 and SQL.

7. Implemented RESTful API web services with Swagger and Java to provide programmatical access to OncoKB data.

8. Developed several microservices by Python, Java, Docker and k8s to implement different features, like keeping the clinic trials data stored in AWS S3 up to date daily, analyzing API resources usage weekly for commercial purpose, and indexing OncoKB private tokens to verify whether they have been exposed on internet.

1. Developed a hierarchical pipeline on-line server system for predicting mitochondrially localized proteins in plants. Its main features are query predicting, task management and result data visualization.

2. Followed RESTful architecture and developed by NodeJS Express framework with creational patterns like factory pattern and singleton pattern.

3. Used Bootstrap 4, CSS 3, HTML to implement website’s pages and used JavaScript to implement responsive interactions between users and website like query uploading. Used D3.js to implement data visualization, including bar chart, line chart and users map.

4. Data of tasks, users and files were stored by MongoDB to implement task management feature with NodeJS.

5. Implemented web crawlers by Python to get results of other on-line tools to be compared.

6. Developed the server by Docker and Apache. The address of the system is: http://mu-loc.org/.

1. Deployed two extensions of a USDA project LANDIS-II Landscape Change Model. These extensions simulate change as forests are influenced by disturbance(fire) and forest management (human harvest). Merged different versions of LANDIS PRO which is old version of LANDIS-II and optimized it.

2. Developed based on .NET Core with factory pattern.

3. Used C# for LANDIS-II on > 50k total code lines by designing and implementing related functions and algorithms, like succession process, harvest events, areas management, sites rank and so on. Used C++ for LANDIS PRO to delete useless codes, merge redundant functions, and add multithread execution.

4. Two models of LANDIS-II could be tested by input image data from 14KB to 4GB within 1h. The execution speed of optimized LANIDS PRO was increased by 30% while the amount of codes was decreased by 15%.