Personalize consumers' health care experience by artificial intelligence!

Building a great Data and Analytics team to grow competitive advantage, capture significant savings, and vastly improve the consumer experience at Cambia Health Solutions.

Built a strong Data Science and Business Intelligence team that reduced CNP fraud at Expedia Inc. by 30% within a year.

Started as a senior scientist and grew to an executive within a year. Led and built a team of data scientists, engineers, and analysts focusing on record linkage, data cleaning and normalization, information retrieval, and information extraction problems at a massive scale.• People record linkage: built a big data entity engine on top of Hadoop that links billions of people records to a core set of 400 million high quality US profiles. • Business record linkage: built a Hadoop-based pipeline that links millions of local business records into a core set of 40 million US business profiles. • Address parsing and normalization using Hidden Markov Model. The system runs on top of Hadoop and can clean and normalize 10 billion addresses in a day. • Person name parsing and normalization using Conditional random fields. • Information extraction from unstructured data: -- Interest extraction of users from their Facebook likes. The interest graph product of LiveFamily app serves 2.5M+ consumers; -- Biographical information extraction from news articles. The system extracted 200M people profiles from local news articles.• Search and ranking: ranking based on purchase and click behavior modeling. • Data cleaning and normalization: mostly machine learning based approaches using class-based language model and logistic regression model.

Conduct research on Information Extraction

Conducted research on web people search.