Episode 17 - MLeap interview: Productionising Data Science

Part one of a two-part interview with Hollin Wilkins and Mikhail Semeniuk, 'the driving forces behind the MLeap project', on why trained Spark ML pipelines are so hard to put into production. Direct audio: https://media.anset.org/roaringelephant/RoaringElephant_Episode_17.mp3