Back

Speaker "Ravi Kumar Meduri" Details Back

 

Topic

A real time processing capability to Sense, Analyze and Connect with shoppers to provide real-time and contextual shopping experience

Abstract

Our customer (retailer) required a solution to ingest high-volume shopper presence events in the store and to correlate and learn from past shopper preferences to offer promotions and useful information to the shopper on the products in real-time. The solution provided capability to process thousands to millions of events per second and compute relevancy using predictive models over terabytes of data. Technologies used were Hadoop stack with Flume, HBase, R-Hadoop and PostgreSQL

Profile

He has more than 15 years of experience in telecom, operations and analytics. Has a strong analytical mind and is experienced with modelling, analytics and big data technologies. He was the leader for an analytics start up where he developed patentable algorithms for customers in e-commerce, digital marketing, and credit and risk assessment in financial services. He works on the entire data lifecycle from collection (extraction) → processing (in-memory) → persistence (data lakes) → transformations → reporting → data mining → predictive and prescriptive analytics (with machine learning). He enjoys most outdoor sports, is an amateur contract bridge player and spends time creating and solving puzzles. He has a few puzzles published with New York Times mathematics blog named Numberplay. Lives in Hyderabad – the city of not just pearls but of warm hearts and food connoisseurs with wife, parents and two children In summary, he has a huge passion for data and analytics and works with evangelical fervour and energy.