Working with the Machine Learning team in YouTube Ads Quality.

Worked with the Creatives team in Search Ads Quality.

Worked with the Creatives team in Search Ads Quality.

• Developed back end ad server code in C++ and Python for a scalable ad experimentation platform in AdWords.

• Developed a map reduce pipeline to automate the extraction of structured information from ad landing pages.

Worked with the Data Mining Group led by Dr. Aristides Gionis on solving the problem of Event Detection in Time Varying Networks.

• Modeled the problem of event detection in temporal networks, where time series data is available for node activity.

• Designed randomized approximation algorithms to solve the NP-hard problem resulting from the modeling, giving proofs for performance guarantees and time complexities.

• Implemented the algorithms in Python and tested their performance over multiple datasets, for empirical validation of the proposed model.

Worked with Dr. Kristina Lerman at the Intelligent Systems Division of the Information Sciences Institute on understanding the Structural Properties of Ego Networks.

• Theoretically derived the relationships between the global and local level structural attributes of social networks, studying degree distribution, degree assortativity, and clustering coefficient.

• Empirically analyzed large real world datasets to validate the theoretical claims, using Python's SNAP graph mining library.