NETWORK SECURITY

Fast and accurate anomaly detection, with no expertise required

The ShieldIOT AI-based network anomaly detection solution analyzes device communications to identify the first signs of cyber-attacks and operational anomalies.  Deployed at the IoT network operator / service provider cloud, the agent-less solution is highly scalable, protecting massive multi-application IoT networks against various threats including DDoS, unknown malware, ransomware, communication hijacking, node takeover and replication, replay attacks, data falsification and many more.

FEATURE HIGHLIGHTS

Unmatched Accuracy

ShieldIOT Corset-based AI anomaly detection is x5 to x20 more accurate compared to leading industry solutions

Real-Time Detection

From one week to one minute. No speed or accuracy trade-offs. Detect the first signs of attack before damage takes place.​

Actionable Alerts

Attack type, duration, affected devices, score, severity, trigger features… Enabling automated or non-expert threat mitigation

Unlimited Scalabilty

Transforms huge volumes of big data into small data sets. Agent-less solution with no device or network changes

ACCURATE ANOMALY DETECTION

The ShieldIOT AI network security is based on coreset real time streaming anomaly detection. The solution analyzes network traffic and additional data sources, to detect the first signs of cyber-attacks,  while providing unmatched out-of-the-box event forensic to facilitate threat mitigation actions.

ShieldIOT is compared to other anomaly detection vendors

ShieldIOT reduces detection time from one week to

Actionable real time alerts and automated threat mitigation with no expertise required

HOW IT WORKS !

The ShieldIOT AI network security is based on coreset real time streaming anomaly detection. The solution analyzes network traffic and additional data sources, to detect the first signs of cyber-attacks,  while providing unmatched out-of-the-box event forensic to facilitate threat mitigation actions.

Data Processing

Data filtering, null completion, data sources joining, data enrichment, feature engineering

AI Engine

Coreset computing to reduce the data, followed by running a set of AI and machine learning detection models.

Anomaly Classification

Automatically generate event type, by combining patented labeling process with domain expertise