Case Study: Cybraics achieves under 5% false positives and faster threat detection with Cloudera Enterprise

Cybraics Harnessing AI to Change the Economics of Cybersecurity

Cybraics’ nLighten platform applies machine learning and AI to detect previously unknown cyber threats, but it faced major scaling and accuracy challenges: exploding security data volumes (tens of terabytes to petabytes), fragmented point tools, extremely high false-positive rates (as high as 95%), and slow analytics that took hours—yet it needed to be delivered as an easy-to-deploy service for enterprise customers.

By running its “Analytical Pluralism” of unsupervised and supervised algorithms and its AI analyst Janus on Cloudera’s modern, distributed data platform (with Apache Spark, Apache Spot and Intel-based compute), Cybraics now processes petabytes quickly, shrinks individual analytics from eight hours to minutes and full-data processing to under an hour, and cuts incident false positives to below 5%, dramatically reducing triage effort and speeding detection and response across industries.

Cybraics

Richard Lovelace

Senior Vice President, Business Development