Evening of technical talks on Apache Flink® in production
We're bringing together data and platform engineers in Amsterdam for an evening of technical talks on Apache Flink® in production.
This edition brings together engineers from Adyen, Booking.com, and Ververica to share real production experiences from powering one of the world's largest payment platforms to operating modern streaming data architectures at scale.
Whether you're already running Flink in production or exploring event-driven systems for the first time, expect practical engineering stories, honest discussions about the challenges behind the technology, and plenty of opportunities to connect with the local data community.
📅 22 October 2026
🕕 6:00 PM – 9:00 PM CEST
📍 Adyen Amsterdam Rokin 49, 1012 KE Amsterdam, Netherlands
What's on
Accelerating and Protecting Shopper Payments with Apache Flink at Adyen
Ayan Banerjee, Thomas van Putten & Vitalii Zhebrakovskyi — Adyen
Data Scientists at Adyen have for a long time developed and continuously improved Machine Learning (ML) models that are utilised at various stages of payment processing. Whether it's fraud detection, payment scheme routing optimisation, bot attack detection, or payouts, ML has consistently delivered better results than traditional rule-based systems and has become a core driver of growth for the business.
With the growing adoption of machine learning, the demand for real-time model features has steadily increased. Custom-built features refreshed daily were no longer sufficient, making reliable, easy-to-build real-time features at global scale a necessity.
In this session, engineers from Adyen's Feature Platform and Fraud Detection teams will share how Apache Flink sits at the core of Adyen's payment processing engine, enabling scalable feature computation that powers millions of transactions every day.
Every Checkpoint is a File: Flink, Iceberg, and the Cost of Freshness
Pedro Mázala — Ververica
Flink and Iceberg fit together cleanly on paper. In production, every checkpoint creates new files, new snapshots, and new operational trade-offs that directly affect query performance and storage efficiency.
Pedro will walk through the practical realities of running Flink with Iceberg at scale, covering topics such as checkpoint frequency, small files, compaction, snapshot management, sink parallelism, and the balance between data freshness and long-term performance. He'll also look at how emerging streaming storage technologies such as Apache Fluss are changing the way these architectures are designed.
Eugene Molokov — Senior Software Engineer, Booking.com
We're excited to welcome Eugene Molokov from Booking.com to share a production engineering story from one of the world's largest travel platforms.
The talk title and abstract will be announced soon.
Who should come
Data engineers, platform engineers, software engineers, architects, and engineering leaders interested in stream processing, Apache Flink, Apache Iceberg, event-driven systems, real-time machine learning, or modern data platforms.
Whether you're running streaming workloads in production or simply curious about where the ecosystem is heading, you'll leave with practical insights from engineers building and operating these systems at scale.
Food, drinks, and networking before and after the talks. Spots are limited, so grab yours early.