Perj
Perj provides file integrity monitoring for cloud security teams that automatically discovers critical files, reconstructs process lineage behind each change, and delivers AI-investigated verdicts with confidence scores. The platform maps findings to MITRE ATT&CK techniques and presents event chains showing the sequence of actions, enabling rapid comprehension of attack paths like SSH persistence and root shell activity.
- Cybersecurity
- Artificial Intelligence
- Software Only
Funding
Founders
Founder details are not available yet.
Product
Problem
Cloud security teams face a high volume of file integrity monitoring alerts that require manual triage to determine whether a change is benign or malicious. Investigating these alerts demands correlating file changes with process lineage and user identity, a time-consuming process that often misses critical context needed to assess actual risk.
Solution
Perj delivers file integrity monitoring that self-configures and investigates its own alerts. The platform automatically discovers critical files on each node and alerts only on relevant changes. When an event occurs, Perj reconstructs the full process lineage and user identity behind it, then uses AI to return a verdict with confidence scores, affected paths, and supporting evidence. The system maps findings to MITRE ATT&CK techniques and presents an event chain showing the sequence of actions, enabling rapid comprehension of attack paths like SSH persistence and root shell activity.
Target Audience
Perj targets cloud security teams responsible for monitoring containerized and Kubernetes-based infrastructure, particularly those focused on detecting file-based persistence and unauthorized access.
Features
- Autonomous file discovery and monitoring across nodes with selective alerting on policy-significant changes
- Process lineage reconstruction identifying the exact executable, shell, and user identity behind each file event
- AI-generated verdicts with confidence percentages and natural-language reasoning for each incident
- MITRE ATT&CK technique mapping (e.g., T1059.004, T1098.004, T1609) for standardized threat classification
- Event chain visualization showing the sequence from container runtime process to file modification
- Indicator-level detail including severity ratings and direct references to source events