Skip to content Skip to navigation
Seminar
10/6/2014 12:00 pm
CBIM Multipurpose Room ( Room 22 )

PageRank-type Algorithm to Determine Creditworthiness

Mr. Vinayak Javaly, Lenddo

Organizer(s): Tina Eliassi-Rad

Abstract

We have created a creditworthiness algorithm, modeled on PageRank, using our members social data. As our cumulative social graph has grown, we have significantly changed our system infrastructure to address scalability issues.

Bio

Vinayak Javaly is an expert in Big Data with over 30 years experience in software development and databases working at IBM, Merrill Lynch, Franklin Templeton and numerous startups. He has worked across the spectrum of computing environments, ranging from UNIX RDBMS multi-tier systems to IBM mainframe systems, and with large to very large databases. During the last fifteen years he has focused on UNIX SQL & NoSQL, datacenter and cloud environments, performing systems design, data modeling, application development, system and database administration, performance tuning and capacity planning. He has built many data warehouses with ETL processes and reporting infrastructure, and designed and managed queueing environments to scale to hundreds of millions to billions of messages per day. Vinayak holds BS & MS degrees in computer science and has been an adjunct professor at NYIT for over 10 years, teaching programming and database courses. He is also a frequent guest lecturer on Big Data at Rutgers University, University of Antwerp, General Assembly, and meetups.