Event Streaming at Scale: Flink at King
Join Data & AI Stockholm, King, and the Apache Flink community for an evening focused on real-time data processing at massive scale, from Flink’s evolving state architecture to how King processes hundreds of billions of game events every week.
This meetup brings together Flink experts and King engineers to share practical insights, real production challenges, and lessons learned from building and operating large-scale streaming systems.
Whether you work with streaming data, event-driven architectures, or real-time analytics, this event offers a rare look into how these systems are designed and run in practice.
🎤 Talks & Speakers
Evolving State Management in Flink 2.0
Apache Flink 2.0 marks a fundamental shift in how state is managed. By decoupling state from operators, Flink evolves toward faster recovery, elastic scaling, and cloud-native execution, while preserving the semantics, guarantees, and programming abstractions that made it successful.
This session explores what changed, what intentionally stayed the same, and how Flink’s state architecture has grown without breaking its core design.
Kingestor: Real-Time Ingestion of Hundreds of Billions of Game Events for ML and Analytics
King ingests and analyzes close to a trillion game events per week, powering game observability, feature delivery, and machine learning for over 200 million monthly active users.
This talk dives into Kingestor, King’s in-house ingestion platform that processes events in near real-time with just ten minutes of latency, handling around five million events per second,while ensuring data integrity through reconciliation and deduplication.
RBEA: Rule-Based Event Aggregation at King
Reza Sadraei
To make stream processing accessible beyond Flink experts, King built RBEA, an internal platform for real-time analytics on live event streams.
RBEA allows teams to define streaming logic using simple scripts, which are executed on production Flink clusters, without managing jobs, infrastructure, or boilerplate. Built on Kafka and Flink, RBEA turns live events into actionable insights within seconds, with results available in BigQuery within minutes.
🗓️ Event Schedule
17:30 – 18:00 Arrival & mingling
18:00 – 18:20 First presentation
18:20 – 18:30 Q&A
18:30 – 18:40 Second presentation
18:40 – 18:50 Q&A
18:50 – 19:00 Third presentation
19:00 – 19:10 Q&A
19:15 – 19:45 Group discussion
19:45 – 20:00 Summary of key discussion points
20:00-20:30 Mingling & goodnight
👥 Who Should Attend?
This event is ideal for professionals working with streaming data, real-time systems, or event-driven architectures, especially those with some years of experience in data engineering, analytics, or platform development.
Expect deep technical discussions, real-world examples, and plenty of opportunities to exchange ideas with peers.
By registering for this event, you agree to the Data & AI Stockholm Event Terms & Acknowledgements.