From Petabytes to Proactive Defense: Powering an AI Swarm With Databricks
Overview
| Experience | In Person |
|---|---|
| Track | Cybersecurity |
| Industry | Enterprise Technology |
| Technologies | Databricks SQL, Lakeflow, Unity Catalog |
| Skill Level | Intermediate |
As AI accelerates cyber offense in a post-Mythos world, the constraint in security operations is no longer data — it’s the speed of the OODA loop. At Arctic Wolf, we process trillions of events daily across 10,000+ organizations, where delayed access to data means delayed defence.
We re-architected our platform on the Databricks Data Intelligence Platform — combining Delta Lake, Unity Catalog, Liquid Clustering, and AI/Genie — to enable real-time, agentic security operations at petabyte scale. Liquid Clustering and Predictive Optimization eliminated data skew and costly acceleration layers, reducing file counts by 50%, improving freshness from hours to minutes, and accelerating 90-day queries by up to 8× (51 seconds to 6.6 seconds).
Architecture: Multi-PB Delta Lake → Liquid Clustering + Predictive Optimization → GOLD telemetry → Genie-orchestrated Swarm of Experts.Outcomes: Sub-second to seconds-level queries, 10–100x alert reduction, >70% precision gains, and earlier detection of cross-tenant attack campaigns.
Attendees will learn how to unify data and AI to compress the OODA loop, power agentic workflows and transform petabyte-scale telemetry into proactive defence.
Session Speakers
Rajneesh Arora
/DSA
Databricks
Justin Lai
/Distinguished Developer
Arctic Wolf