Keynote

10:05 AM - 10:35 AM PST , October 25

Streaming Machine Learning with Flink, Pulsar & Iceberg

Discord is the place to talk online, whether that’s one-on-one, in small groups, or in larger communities organized around shared interests. In this talk, we'll show how Discord uses Apache Flink to power real-time machine learning applications for fighting abuse at scale & keeping over 150M active users safe. We'll share the how and why of our migration to Pulsar from Google Pub/Sub, and how we pair Pulsar with Apache Iceberg to create a data layer capable of seamless historical and realtime serving. Together, the three technologies unlock faster feature engineering, backfilling, point-in-time accuracy, and minimize offline-online skew, making this architecture compelling for practical real-time ML in production.

Speaker

David Christle

Staff Machine Learning Engineer, Discord Inc.