1. Designed and Implemented an online PCB boards defects detecting system, with a better than human accuracy of more than 95%, supervised by Andrew Ng.

2. Used Flask as the server side framework for neural network computing and implemented an Android app as the client end to take photos and show process results.

1. Developed C# application for computer vision research, data visualization and annotation coupled with feasibility to easily annotate videos and images, by incorporating object matching and tracking algorithms.

2. Developed Python server side functionality for a novel graph based web user interface to present, explore and search structured data based on feature distances between data nodes. The web user interface is based on Django framework with MongoDB database.