Perpetual Systems

Perpetual Systems is a cybersecurity company that develops a full-stack detection and response platform designed for AI-scale security operations. The platform is built around a security data lake architecture that runs within the customer's own cloud environment, using open formats (Iceberg/Parquet) to store and query security data without per-GB ingest pricing markups. Core platform capabilities include: - Agentic Detection & Response: An AI agent named Simba orchestrates a set of specialized sub-agents to handle threat intelligence ingestion, alert triage, false-positive resolution, and case escalation to human analysts. - Threat Scenario Detection: Correlates security signals into end-to-end threat actor behavior patterns to reduce noise and surface high-value alerts. - Query Engine: A proprietary query language called Hamelin supports detection engineering, threat hunting, aggregation, filtering, sliding window correlations, and behavioral baseline comparisons. - Semantic Layer: Security events are normalized and cataloged, allowing the AI agent to reason over security concepts rather than raw database tables. - Security Data Lake: Customer-owned storage based on Iceberg/Parquet, with compute and storage costs paid directly by the customer. The platform operates as a continuous analytics flywheel: detections are suggested from threat intelligence, alerts are investigated by agents, outcomes feed back into detection tuning, and low-value detections are flagged for improvement. This loop is intended to compound SOC efficiency over time. The company was founded by Christian Beedgen (previously co-founder and CTO of Sumo Logic, and early engineer at ArcSight) and Kyle Derr (previously Chief Architect at Sumo Logic and Jask Labs). The team's background spans over 25 years of SIEM and log management development.