Interview
summary
Building GoPro's data platform for four years, creates a Spark-Kafka streaming ingestion pipeline that pushes device events into storage and near‑real‑time dashboards, then layers on tooling to become a machine learning platform for collaborative model work. Describes pragmatic best practices learned from failed attempts, a recurring pain point around table management, and a migration from Java to Scala because functional patterns map naturally to Big Data workloads.
AI-written summary · about
related talks
04minterview
26mRe-architecting the Data Platform with Spark on Kubernetes
44mProcess and analyze millions of events per second with open source tools
40mProductionizing Big Data - Stories from the Trenches
30mBeyond the Build: Unleashing the Hidden Value in Your Data Platforms
33mAccidental DataOps: How we matured our organization with data, and how you can too!
46mKafka Streams at the Big Data Crossroads at Lightning Speed
54mData engineering & DataOps: Ask Me Anything