Apache Spark, Scala, Reactive Technologies and Machine Learning discussions


Details
Agenda:
6:00pm - 6:30pm - Arrive and socialize
6:30pm - 7:00pm - Discussion 1: Building a Reactive RESTful API with Akka HTTP and Slick
7:00pm - 7:30pm - Discussion 2: Cloud Scale Machine Learning with ApacheSpark, MLLIB, Cloudfoundry and OpenStack Swift - Romeo Kienzler - Chief Data Scientist - IBM Cloud Transformation Lab Zurich
7:30pm - 8:00pm - Discussion 3: The Typesafe Reactive Platform and Apache Spark: Experiences, Challenges and Roadmaps - Stavros Kontopoulos, Typesafe
Discussion 1
Title: Building a Reactive RESTful API with Akka HTTP and Slick
Abstract:
Zalando’s engineering team is currently rebuilding our “shop”—the unit that includes our 15 country-specific, customer-facing websites—to transform it from a monolith into a microservices architecture. As part of this work, we have developed open-source tools in a variety of languages. One of this tools, Innkeeper, is a simple, RESTful route management API built on top of reactive technologies like Akka HTTP and Slick. It will help our fleet of router instances to keep their routes in sync.
In this presentation, I'll discuss Innkeeper’s architecture and how — by choosing a reactive stack of technologies — we can stream database records back to the browser. I’ll talk about how (and why) we secured our API using OAuth, and give insights on how we use Docker and the STUPS tools to auto-scale this API on top of AWS. Finally, I’ll describe how we designed a set of unit and integration tests to make sure that our application works correctly, and how we set up our continuous integration environment.
About the speaker: Dan Persa has been a software engineer at Zalando since 2013 and is a member of the Fashion Store team, which is responsible for Zalando’s core ecommerce business. He loves Java and Scala and more recently has been exploring Go and Node.js. He’s a big fan of clean code and software craftsmanship. In addition to coding, he enjoys mentoring new developers, organizing coder dojos and reading groups and giving tech talks. In his free time he likes to take photos and dance salsa.
Discussion 2
Title: Cloud Scale Machine Learning with ApacheSpark, MLLIB, Cloudfoundry and OpenStack Swift
Speaker: Romeo Kienzler - Chief Data Scientist - IBM Cloud Transformation Lab Zurich
Abstract: Clouds provide very cost effective access to computing power and storage. Especially on workloads with high variation on computational and storage resource requirements. Resilient data processing engines like Apache Spark, a parallel machine learning library on top like MLLIB and elastic storage services like OpenStack Swift are a perfect fit to build a Resilient data processing framework. And this is exactly what IBM has done on the IBM Platform as a Service Cloud called Bluemix. In this session I'll give a very short introduction into all these technologies, open standards and frameworks in order to prepare you for a end - to -end demonstration of a large DNA sequence analysis.
About the speaker: Romeo Kienzler holds a Master Degree in Information Systems with specialisation in Applied Statistics and Bioinformatics from the Swiss Federal Institute of Technology. He works for IBM Zurich as a chief data scientist and architect. His current research focus is on large-scale machine learning on resilient cloud infrastructures based on Apache Spark and Apache Flink on top of Docker Swarm.Romeo Kienzler is a Member of the IBM Academy of Technology, the IBM Technical Expert Council and the IBM BigData BlackBelts team.
Discussion 3
Title: The Typesafe Reactive Platform and Apache Spark: Experiences, Challenges and Roadmaps
Speaker: Stavros Kontopoulos, Typesafe
Abstract: In this presentation Stavros Kontopoulos will discuss the role Apache Spark plays for the Typesafe Reactive Platform. He will discuss the Spark team's contribution to the open source project, cover added features like backpressure and dynamic allocation, and describe challenges the team has faced while developing infrastructure for integrated testing. He will also share upcoming roadmap plans, such as:
kerberos security, already there for hadoop a must for the enterprise world.
Closely examine Tachyon technology.
More closely integrate Spark with Typesafe Reactive Platform and our commercial product ConductR
About the speaker: Stavros Kontopoulos recently joined the Apache spark team at Typesafe. Prior to Typesafe, Stavros worked in telecoms for several years developing network management products. Immediately before Typesafe he worked as a big-data engineer in the digital marketing sector for a start-up (Pollfish Inc). Stavros has extensive hands-on experience with web technologies. When not coding he is doing research pursuing a Phd degree with areas of interest being: sublinear algorithms, big data estimators, massive data indexing structures, social network analysis and data structures.


Sponsors
Apache Spark, Scala, Reactive Technologies and Machine Learning discussions