[Colloq] Colloquium Talk: Friday, December 6 @ 11 am in 366 WVH: Mayur Thakur "A Tale of 3 Data Types: Surveillance Algorithms over Text, Graph, and Numeric Data/ Faculty Host: Ravi Sundaram

[Khoury Academic Affairs]

Colloquium Talk: Friday, December 6 @ 11 am in 366 WVH: Mayur Thakur "A Tale of 3 Data Types: Surveillance Algorithms over Text, Graph, and Numeric Data

Date: Friday, December 6, 2019

Speaker: Mayur Thakur

Talk Title: "A Tale of 3 Data Types: Surveillance Algorithms over Text, Graph, and Numeric Data"

Time: 11:00 am - 11:50 am

Location: 366 WVH


Abstract:

Each day petabytes of financial data flow through the pipes of large financial institutions such as Goldman Sachs. These include billions of market events, millions of trades, emails, and deals. From billions of little pieces of data, algorithms mine patterns and find a relatively tiny number of risky events such as insider trading, market manipulation, and leakage of confidential information.



In this talk we will we will show that building robust surveillances requires solving, at scale, some well-studied problems in distributed systems, search engines, NLP, and graph algorithms. For example, non-parametric outlier detection and k-nearest neighbor techniques are crucial in detecting money laundering.


Bio:


Mayur Thakur is the head of Surveillance Engineering in the Global Compliance Division. He joined Goldman Sachs as a managing director in 2014. In the past 5 years, he has led the development of a big data surveillance engineering platform, which now handles petabytes of data and runs a variety of surveillances. In his capacity as the head of surveillance engineering, Mayur has been responsible for building a team of engineers who have skills in big data, modeling, and finance. Prior to joining the firm, Mayur worked at Google, where he designed search algorithms for more than seven years. Previously, he was an assistant professor of computer science at the University of Missouri.



Faculty host: Ravi Sundaram