Processing 100 Billion events a day at GumGum
Details
Abstract:
Being deeply integrated within the Ad Tech ecosystem, GumGum participates in tens of billions of real-time auctions every day. As a result, they're processing and reporting on over a 100 Billion events a day. During this presentation, Maxime Nay will explain how GumGum's Data Engineering team built a scalable and highly-available real-time reporting platform. He'll go over the current architecture of their systems, the roadblocks that GumGum had to tackle in order to reach this scale, and what are the current challenges they're facing. GumGum's data platform runs a 100% on AWS and uses many major open source Big Data tools, including Spark, Kafka, Druid etc.
Speaker Bio:
Maxime Nay is the Tech Lead of GumGum's Data Engineering team. He's been with the team for five years and has worked on a multitude of data oriented backend applications. While at GumGum he worked with numerous Big Data technologies such as Spark, Kafka, Storm, Druid, Cassandra and has extensive experience running those technologies on AWS. Maxime is originally from Lyon, France and moved to Santa Monica in 2013.
Processing 100 Billion events a day at GumGum