Protecting the heartbeat of ML/AI use cases with HYCU

HYCU’s presentation at Cloud Field Day focused on the critical need for data protection within the rapidly expanding AI/ML landscape. The increasing adoption of AI mandates across organizations necessitates robust protection for the underlying data lakes and lake houses that fuel these systems, as well as the repositories AI creates. The presentation highlighted the broad coverage HYCU provides for this data stack, emphasizing its innovative solutions for Google BigQuery, a major data framework used in production environments.

A key aspect of the presentation centered on the various reasons why protecting AI data is essential beyond simply recreating results. Speakers discussed the importance of cyber resilience, the ability to revert to specific points in time to address performance issues or model drift (re-vectoring), and the crucial role of data protection in meeting legal and compliance requirements, such as demonstrating the absence of PII or IP infringement in training data. Furthermore, the complexity of reconstructing datasets spread across diverse sources (on-premises and cloud) was underscored as a significant challenge requiring a comprehensive data protection strategy.

The presentation showcased HYCU’s capabilities in addressing these challenges, specifically demonstrating its solutions for BigQuery. HYCU’s platform provides automated discovery and protection for a wide range of Google Cloud services and boasts a patent-pending technology enabling atomic backups. This innovation addresses the critical issue of data synchronization across multiple tables and datasets within a data lake house, ensuring consistency during backups and recovery. The discussion also highlighted the increasing reliance on data lake houses as central repositories for AI-related data, emphasizing the importance of robust protection for these often singular copies of crucial datasets.