Data: Moving with the flow

on YouTube ↗25m

" summary

Argues that stream processing replaces slow batch jobs with near‑real‑time pipelines using Apache Flink; cites Alibaba running Flink at 19–20,000 CPU cores analyzing ~16,000 behavioral data points for per‑user recommendations, Netflix recomputes device‑level suggestions on each interaction, and banks use Flink for sub‑second fraud checks and real‑time transfers. Claims Flink apps (not jobs) port across cloud, hybrid and edge, preserving consistent business logic and meeting GDPR, latency and scale constraints.

AI-written summary · about

" chapters (12)

1. 0:00Start
2. 1:26How Data Is Processed
3. 2:18Stream Processing
4. 3:13Real-Time Processing
5. 4:00Why Is It Processed
6. 5:23Private and Public Cloud
7. 5:43Geographical Distribution
8. 7:14Maintainability
9. 14:10Netflix
10. 17:10Real-Time Money Transfers
11. 17:49Credit Card Fraud Detection
12. 18:32Pinterest

" related talks

• 49mscale.bythebay.io: Stephan Ewen, Apache Flink and the Next Wave of Stream Processing ApplicationsStephan Ewen · Scale By the Bay 2017
• 45mApache Flink — Real-Time Data ProcessingGaël Renoux · Devoxx France 2024
• 47mBring Flink to Large Scale ProductionXiaowei Jiang · Flink Forward Berlin 2017
• 34mBeaming Flink to the Cloud @ NetflixMonal Daxini · Flink Forward Berlin 2016
• 35mApache Flink 2.0: Unified Enterprise Data Processing System and BeyondBowen Li · Scale By The Bay 2019
• 41mBeyond the Hype: When Apache Flink Solves Real ProblemsNaci Simsek · Berlin Buzzwords 2026
• 42mStream Processing beyond Streaming DataStephan Ewen · ITNEXT Summit 2019
• 43mLeverage the power and simplicity of SQL on Apache FlinkFabian Hueske · Berlin Buzzwords 2018