Real-time Customer Engagement in Gaming Using Kafka and Flink
summary
Concrete claims and runnable details guide the choice here: a production pipeline built with Kafka and Flink handles 1.5 billion daily events, uses a broadcast state pattern to push JSON campaign rules and JavaScript filters at runtime, evaluates campaigns and segments in-stream, writes triggers back to Kafka, and achieves subsecond end-to-end latency. The code and a demo UI are open source and the live demo shows the pipeline working end to end.
AI-written summary · about
chapters (9)
- 0:00Stream processing in gaming
- 3:56Innog Games architecture
- 7:12Kafka basics
- 12:54Flink fundamentals
- 19:57Real-time engagement use case
- 26:22Broadcast State pattern
- 30:14Live system demo
- 36:57Future outlook with machine learning
- 38:43Key takeaways
related talks
40mReal-time Customer Engagement in Gaming Using Kafka and Flink
42mStream Processing with Apache Flink
47mReal-time Fraud Detection Challenges and Solutions
51mStream Processing Smackdown: Kafka Streams vs. Apache Flink
20mReal-time analytics with Flink and Druid #bbuzz
41mStream Processor as a Database: Building Online Applications
38mHow to handle 100k events per second
47mFlowable business processing from Kafka events