DESG x Confluent x IMDA Nov 2026

Hi DESG Community,

It's been a while since we last met and for this round, we have Weicong from Confluent and Vivek from Julius Baer who will respectively talk about:

Kafka Producer Performance Tuning: A Metrics-Driven Approach
"In this session, we'll start with default producer configs and progressively tune batch.size, linger.ms, and compression through live demos — watching key metrics like bufferpool-wait-ratio and record-queue-time-avg change in real time. You'll learn how these parameters interact inside the producer's buffer, why tuning order matters, and how to read the metrics output."

AI Agents Can't Wait for Batch: Building the Real-Time Streamhouse with Apache Flink & Fluss
"AI agents are only as intelligent as the data they rely on. Discover how Apache Kafka, Apache Flink, and Apache Fluss complement each other in building the next-generation Streamhouse, connecting real-time streaming, continuously updated state, and historical analytics.
Through a practical payment-risk demo, we'll explore how this architecture enables AI agents to make smarter decisions using fresh, reliable data, while examining the evolving role of streaming technologies in modern data engineering."

Expected agenda:
6.30pm - 7.15pm - Registration\, Pizzas & Networking
7.15pm - 7.20pm - Introduction
7.20pm - 8:00pm - Talk by Weicong
8.00pm - 8:40pm - Talk by Vivek
8.40pm – 9:00pm - Q&A
9.00pm – End of event

Hope to you all there!

Cheers!
DESG