Skip to content

Details

PubMatic serves billions of advertising impressions and collects terabytes of data from several data centers across the world. In batch data processing, data is collected at different geographic locations and processed at regular intervals. This system brings delay of at-least 1 hour before an event is accounted for.

The goal of having realtime streaming @ PubMatic was to provide Publishers, Demand Side Platforms (DSPs) & Agencies actionable insights in a few minutes from the time of event generation

PubMatic uses DataTorrent RTS powered by Apex for

Real time reporting.

Resource monitoring.

Real time learning.

Allocation Engine.

Dev Tagare from the PubMatic team will take the audience through the architecture, custom operators developed, use-cases for realtime & the challenges involved in implementing streaming systems at scale where multiple data centers are in play.

Dev is a Data Architect with the BigData platform group @ PubMatic & have been working actively on DT-RTS (powered by Apache Apex) since May 2014. Dev will be supported by Thomas Weise from Data Torrent. Thomas is principal architect at DataTorrent and has developed and architected distributed systems, middleware and web applications since 1997. Thomas joined DataTorrent at its inception

Agenda

6:00-6:15 Pizza, networking

6:15-7:30 PubMatic use of DataTorrent/ Apex

7:30-7:40 Discussion/Q&A

Related topics

You may also like