Loves Data!
I currently work as a Senior Modeling Scientist in the Publisher Engineering team at Quantcast. My team is responsible for Quantcast Measure. My primary interests are in algorithms and data mining in relational domains.
I grew up in India and finished my initial schooling at Smt. Sulochanadevi Singhania School. In 2007, after my undergraduate studies at the University of Mumbai, I came to the United States to pursue graduate education. I hold a PhD degree in Computer Science from Stony Brook University and was advised by Prof. Jie Gao.
My work at Quantcast involves applying inference techniques to large volumes of data that power Quantcast’s measurement products. I have worked on performing demographic, unique visitor and interest modeling as lead scientist on the Publisher Engineering Team at Quantcast.
My doctoral research focused on analyzing dynamics in online social networks. I worked on the ADAMS project. In particular, I worked on modeling and predicting departure events in social networks and detecting social bias on online consumer review websites. I have previously also worked on large-scale time-series and sentiment analysis for news sources as member of the Lydia project.
My research and work has been covered by various media sources.
I played a major role in conceptualizing Audience Interests and Audience Grid products for Quantcast Measure.
I designed the indexing scheme that is currently used to rank the results. I also implemented the backend pipeline that powers these products.
I played a pioneering role in modeling and implementing the Political Affiliation and Engagement demographic segments for Quantcast Measure.
These segments help capture the political affiliation and degree of political engagement for a web property's US audience. I implemented and supported the pipeline that produced these segments in a very short period of time (1 month). This allowed Quantcast to be the first free measure tool to provide these insights for all quantified web properties in the US. The product was released (in Beta) in September, 2014 to align with the US Midterm Election Cycle.
We conduct field experiments that provide empirical evidence of Success Breeds Success Dyanamics. The study was published in the Proceedings of the National Academy of Sciences (PNAS).
My Google Scholar profile is available here. My Erdős number is 3. The Erdős number gives the collaboration distance between a person and the mathematician Paul Erdős.
The best way to contact me is by email. This is especially true if you are reaching out to me for the very first time and / or we have never met.
