Breaking the Ice(berg): Riskified’s Journey to its Next-Generation Lakehouse

How many of you manage multiple data stores in your organization, wrestling with different data formats that led to data duplication, infrastructure redundancy, and uncontrolled costs? Imagine reducing these costs by 25% while maintaining your existing SLAs — this is exactly what we achieved with our next-generation architecture.

In this session, we'll show you how we built Riskified's next-generation lakehouse by leveraging Databricks' native Apache Iceberg support and Unity Catalog. We'll share our innovative approach to cross-platform querying without data duplication, and how we transformed our data warehouse into a modern lakehouse architecture. Throughout the session, we'll explore the technical challenges we conquered, from data migration to performance optimization. The result? A simplified world where everything is identical across engines, leaving users with just one choice — which query engine best suits their use case.