Bloomberg Tech Talk

Bloomberg Tech Talk<br><b><i>From Inception toProduction: How Bloomberg Leverages Open Source Technology to Move Data ScienceExperiments from the Lab to the Real World</i></b>&nbsp; &nbsp;&nbsp;<br>Speaker:&nbsp;Ilan Filonenko, Software Engineer<br>Date: January 15, 2020<br>Time: 5:00pm - 6:00pm<br>Location: Gates Building, Room 219<br><br>RSVP via Bloomberg's link (<a href=""></a>&nbsp;) and Handshake (<a href="">https://stanford.joinh...)<br><br>Abstract:<br>Given Bloomberg's well-established culture of innovation, collaboration, and transparency, we consider open source software to be a valuable element in producing high-quality software. Bloomberg has invested heavily in open source technology at all levels of our tech stack, including infrastructure (Kubernetes), ETL pipelining (Apache Spark), and machine learning (TensorFlow). Bloomberg has hardened these technologies beyond the industry standard to provide production-grade data science tools for its 6,000+ engineers in the form of its internal Data Science Platform. The Data Science Platform provides tenants with secure, reliable, and scalable solutions for their machine learning workflows and ETL pipelines. In this talk, we will discuss two aspects of this platform:how we developed our model development lifecycle (MDLC) and how we productionized Spark in a secure and scalable multi-tenant fashion by promoting adisaggregated architecture.&nbsp; &nbsp;&nbsp;<br><br>Bio:&nbsp;<br><b>Ilan Filonenko</b><br><p><a href=""></a></p><p>Ilan Filonenko is a member of the Data Science Infrastructure team at Bloomberg, where he has designed and implemented distributed systems at both the application and infrastructure level. He is one of the principal contributors to Spark on Kubernetes, primarily focusing on enabling Secure HDFS interaction, non-JVM support, pluggable remote storage for shuffle files, and dynamic allocation capabilities. Ilan’s research has focused on algorithmic, software, and hardware techniques for high-performance machine learning: focusing on optimizing stochastic algorithms and model management.</p>

Wednesday, January 15, 2020 - 5:00pm to 6:00pm
Gates Computer Science, 353 Serra Mall, Stanford, CA 94305, USA