Temesgen Meheret

Enterprises often run data warehouse native tables alongside open formats, which drives duplicate storage, complex governance and higher costs. To address this, an enterprisewide initiative launched to migrate end-user owned Snowflake spaces from native tables to Apache Iceberg tables within a unified cloud object store (S3). This session covers the target architecture, migration and automation patterns, and how defaulting table creation to Iceberg simplified developer adoption. Attendees will learn how this approach reduced vendor lock-in, eliminated redundant storage and delivered measurable cost savings at scale.

As one of Snowflake's earliest and largest adopters, Capital One has continually driven innovation in cloud data management, focusing on optimizing cost and governance at a massive scale. This session details our strategic journey toward a Bring Your Own Compute model, illustrating how we evolved our architecture to meet rigorous security, governance and cost optimization requirements in a federated data environment. A core component of this evolution was implementing a sophisticated architecture that allows data consumers flexibility while maintaining central control via multi-catalog and multi-table format architecture, achieving data residency and compliance, and maximizing finops efficiency.

Open table formats empower companies to bring compute to their data while maintaining a single copy of data. At Capital One, we use Iceberg Tables to optimize ingestion and storage costs in Snowflake without compromising data integrity or performance. This enables us to expand how we leverage Snowflake's capabilities for diverse workloads. In this session, we'll share how Iceberg helps us democratize compute choices along with key learnings from migrating datasets from native Snowflake tables.

In this session, you learn how to bring business intelligence (BI) to everyone in your organization using Amazon Q in QuickSight. Explore connecting different data sources, managing users and groups at scale, and implementing operational best practices.