Druid, Realtime Analytics on Big Data & a Lots more...

This two hour session will focus on frameworks and topics that enable Realtime, scalable analytics on Big data.

Session - 1: Making the Elephant Dance... again! - This talk will focus on challenges and techniques around how to get meaningful insights out of large datasets, Current challenges that exists within PayPal and our various approaches towards this goal.

• Speaker: Subho Ghosh (Director Of Engineering, PayPal)

Session - 2: Druid -(http://druid.io) - Open source infrastructure that provides for exploration of very large quantities of data as it is ingested into the system. Druid was created out of necessity by Metamarkets, a company focused on providing real-time interactive insight to the RTB (real time bidding) AdTech space with a full stack analytics service. Metamarkets required a system that could ingest data in real-time, provide ad-hoc N-dimensional drill down and still provide sub-second responses. As a hosted service, Metamarkets also required no downtime deployments, fault-tolerance and self-healing properties.

• Speaker:Nishant Bangarwa (Software Engineer, MetaMarkets)

Session - 3: Druid @ PayPal - This talk will focus on our efforts to ingest PayPal's clickstream data and provide reporting capabilities to slice and dice data across multiple dimensions and arbitrary levels of drill down.

• Speaker: Suresh Kumar (Engineering Manager , PayPal), Vikram Ramakrishnan (Software Engineer , PayPal)

Session - 4: Fast, Cheap, and 98% Right: Cardinality Estimation for Big Data - The nascent era of big data brings new challenges, which in turn require new tools and algorithms. At Metamarkets, one such challenge focuses on cardinality estimation: efficiently determining the number of distinct elements within a dimension of a large-scale data set. Cardinality estimations have a wide range of applications from monitoring network traffic to data mining. If leveraged correctly, these algorithms can also be used to provide insights into user engagement and growth, via metrics such as “daily active users.”

• Speaker: Nishant Bangarwa (Software Engineer, MetaMarkets)