
Embedded AI

Sensor Operations
• Problem: Critical Windows security events (failed logins, privilege changes, Kerberos errors) often drown in log volume and go unnoticed.
• Solution: EventLog Sentry monitors configured Event IDs, cleans messages, tracks first/last seen, counts, and rates, then enriches with AI anomaly scoring and syslog export.
• Why Different: Embeds ONNX AI models for adaptive thresholds, anomaly detection, and autonomous learning; supports JSON syslog, registry -driven config, and agentic AI updates every 15 minutes.
• Syslog Output Fields: Timestamp, Sitename, Host, Service, Event ID, Flags (e.g., Login Success, User Created, Locked Out), Count, Rate (/min), Duration, Threat Score, Anomaly Flag, Recommended Action.
Host Resource Requirements
-
CPU: < 1%
-
Memory(RAM) < 50MB
-
Disk: < 1MB per day log writes
-
Network: < 1KB/s syslog traffic




