Currently working on Facebook Ads Ranking problems.

Teaching Assistant for 11-442 / 11-642 Search Engine in Language Technologies Institute.

Teaching Assistant for 17-683 Data Structures for Application Programmers in School of Computer Science.

· Built a video surveillance desktop application for property management; Features include video synopsis, object segmentation, motion detection, face recognition and human tracking

· Developed at front-end on the video synopsis and object segmentation feature using Tencent’s proprietary Dynamic UI Lib using C++; Features ready for testing

· Developed at back-end on the human tracking feature using Spring Boot framework using Java

· Scrapped enterprise information from Google Place API and websites; Built a pipeline to automate data preprocessing

· Classified business industry with NLP; Technology includes word embedding with SVM, Random Forest and Neural Nets such as RNN, 1-dimensional CNN combined with LSTM; Trained models and built pipelines from scratch; Improved accuracy by 40% on the baseline (keras, sklearn)

· Segmented customers for the client with unsupervised model, generating meaningful and intuitive explanations