Iceberg v3 in Production: From CDC to Governed AI-Ready Data

Apache Iceberg v3 introduces deletion vectors, row lineage, VARIANT, nanosecond timestamps, and other capabilities built for high-churn, semi-structured, and governed data workloads. But adopting v3 raises real questions about engine compatibility, migration risk, and ongoing maintenance.

This session gives you a practical framework for making that call, and a look at what running v3 in production actually looks like.

You’ll learn how to:

• Compare Iceberg v2 and v3, and decide when migration makes sense for your tables
• Run a production workflow for CDC updates and deletes using v3’s deletion vectors
• Track row-level lineage and apply table maintenance under the new format
• Build a v3 readiness framework so you can adopt new capabilities without unnecessary replatforming

Attendees will leave with a clear v3 readiness framework and practical guidance for building AI-ready data foundations without unnecessary replatforming.