Paris Fast Data Meetup

IMPORTANT PLEASE RSVP: https://www.meetup.com/paris-apache-kafka-meetup/events/313343583/?eventOrigin=group_upcoming_events

***
Rejoignez-nous pour une rencontre Apache Kafka® le mercredi 18 mars à partir de 18h00, organisée par meshIQ et Peerspace!
Les présentations seront en français.

📍Lieu
SHAPES Studio via Peerspace
92 Rue René Boulanger
75010 Paris

🗓 Programme

• 18h00 : Ouverture des portes/Accueil, Boissons
• 18h30 - 19:20h : Gilles Philippart, Software Practice Lead, Confluent
• 19:20h - 20h10 : JB Onofré, Ingénieur logiciel principal, Dremio + Directeur de la fondation Apache
• 20:10h - 20h30 : Restauration, Questions-réponses supplémentaires, Networking

💡 Intervenant 1 :
Gilles Philippart, Software Practice Lead, Confluent

Discussion :
Streaming into Iceberg: Kafka Connect vs Flink vs Tableflow

Résumé :
Getting streaming data into Iceberg tables sounds simple—until you hit schema evolution, upserts and deletes, small file problems, and commit contention. In this talk, I’ll walk through the real-world challenges of streaming ingestion into Iceberg and compare three common approaches: Kafka Connect, Flink’s Iceberg sink, and Confluent Tableflow.

Biographie :
Gilles works at Confluent as a Software Practice Lead and has over twenty years of experience in the financial sector, primarily developing systems for investment banks. In recent times, he served as a Principal Engineer at a FinTech company, where he built a platform that offers loans to small and medium businesses in the blink of an eye. Through this work, he became a proponent of event-driven architectures with Apache Kafka and is now passionate about guiding others to leverage its advantages. Outside of work, Gilles enjoys reading books, whipping up culinary delights, and playing games with his daughter and wife.

💡 Intervenant 2 :
JB Onofré, Ingénieur logiciel principal, Dremio + Directeur de la fondation Apache

Discussion :
From Legacy Queues to Modern Tables: Bridging ActiveMQ, Kafka, and Apache Iceberg

Résumé :
Modern data platforms rarely start with a blank slate. Organizations often face the challenge of unifying "classic" messaging systems like Apache ActiveMQ with high-throughput streams like Apache Kafka to feed a unified analytical layer.

In this session, we explore how to build a robust ingestion pipeline that captures events from diverse sources and commits them into Apache Iceberg. We will dive into:

• The Bridge: Strategies for migrating or mirroring data from JMS-based ActiveMQ to Kafka.
• The Sink: Using Kafka Connect and Iceberg’s native capabilities to handle schema evolution and ACID transactions.
• The Payoff: How this architecture enables "time-travel" queries and high-performance analytics on streaming data without the "small-files" headache.

Biographie : JB est directeur de la fondation Apache. Il est membre de plus de 20 projets Apache et également mentor sur de nombreux podlings. Il est aussi en charge de l'équiipe OSPO chez Dremio.