Title: "The other 80% - Taking Spark Streaming to 24/7 production"
Speaker: Martin Lehner (OVO Energy)
Abstract: When we realized our old architecture for OVO Live - our live usage app - would not take us much farther than the current user count, we decided to switch to a modern big data architecture. With live data, streaming comes quite natural - and in our analysis Spark with both its batch and stream capabilities came out on top. This talk focusses on the challenges we faced actually implementing, deploying, updating and operating a streaming architecture with Spark, and how to tackle them. Tied to a real world example, looking back on 6 months with our new architecture.
Title: From Spreadsheet to Streams.
Speaker: Michael O'Kane (Capgemini)
Abstract: A walkthrough of an event processing architecture deployed with some of our legacy-heavy clients. Using Drools, Spark and yes, even spreadsheets to effectively develop, deliver and govern complex business processes on modern big data architectures (with surprisingly little effort). Covering the how and the why, this talk will explore one effective way we've found to bridge the gap between behoodied Scala wizards and established development capabilities, with a little bit of Spark 2.0 thrown in for good measure.
Bio: Michael is Big Data Engineering Lead at Capgemini UK, primarily engaged in the UK Public Sector.