IronTin Sentinel - AI-Powered Behavioral Anomaly Detection and Threat Prediction

IronTin Sentinel applies machine learning models and behavioral analysis to predict security threats before they manifest. Built with Rust and local ML model inference (no cloud APIs), it collects behavioral baselines -- process lifecycles, network patterns, file access rates -- and runs anomaly scoring against those baselines. Key innovations: (1) PIR (Predictive Incident Response) analysis: a proprietary anomaly scoring pipeline that assigns risk weights to deviations from established behavioral norms; (2) Local ML inference using embedded ONNX models with no data leaving the device; (3) Threat feed integration for cross-referencing anomalies against known attack signatures without cloud reporting. Defends: Anomalies, Behavioral Threats, AI Prediction, PIR Analysis. Part of IronTin Security Suite v1.0 (December 2025).