Junior Web developer for client commercial web development and Maintenance. Supported clients need by setting up websites, creating user accounts and troubleshooting through helpdesk inquiries.

• Migrated static content to Content Management System WordPress.
• Utilized PHP, HTML, CSS, JavaScript and MSSQL to create online applications
• Updated website content with WordPress, HTML/CSS and FTP.
• Maintenanced website through PHP and Dead Link Checker

Junior Web developer for client commercial web development and Maintenance. Supported clients need by setting up websites, creating user accounts and troubleshooting through helpdesk inquiries.

• Migrated static content to Content Management System WordPress.
• Utilized PHP, HTML, CSS, JavaScript and MSSQL to create online applications
• Updated website content with WordPress, HTML/CSS and FTP.
• Maintenanced website through PHP and Dead Link Checker

Freelance front end web designer/developer in startup company named CampusU. CampusU is a social media website which connect all university students together. Provided platform to students to create personal account, chat and upload notes.

• Managed static content through Content Management System WordPress.
• Designed and updated database through Microsoft SQL.
• Polished website user interface through HTML/CSS and JavaScript.
• Controlled website version through Github.

This web application is fundamentally interconnected and peer to peer. used bootstrap to decorate webpages. Users are allowed to create their own account, profile, share their posts and be able to follow other users in order to read their posts.

• Used Python2.7 and web framework Flask developed the twitter-like web application.
• Used PostgreSQL design and develop database.
• Deployed and migrated web application on the Heroku server.

This project is from the competition of Kaggle, our purpose is building the predictive model to distinguish between main product categories of Otto Group products. Firstly, we are using Gradient boosting trees, Gradient boosting logistic regression, Neural network, Random forest and linear discriminant analysis to build stacking models in order to classify a dataset with 93 features for more than 200,000 products. Secondly, we use folds for train dataset to predict the stacking models. Thirdly, we use calibration and model ensemble technique to minimize the logloss. Finally, our lightweight model achieved a logloss of 0.42593, and our team’s rank is in top 8% for this Kaggle competition.