Databricks Announces Lakewatch: New Open, Agentic SIEM

Databricks announces Lakewatch, an open, agentic SIEM built on lakehouse architecture to counter AI-driven cyberattacks at machine scale.

• •Unifies security, IT, and business data in a single governed environment using open formats (OCSF, Delta Lake, Apache Iceberg)
• •Decouples storage from compute to enable petabyte-scale full-fidelity telemetry retention at reduced cost
• •Embeds AI agents (Genie) for automated log ingestion, detection authoring, false-positive reduction, and natural language threat hunting
• •Supports Detection-as-Code via YAML/SQL/Python with CI/CD pipelines and custom ML models through MLflow
• •Powered by Anthropic Claude for cross-data signal correlation; acquiring Antimatter and SiftD.ai to deepen agentic security capabilities