Case Study: Teridion achieves 99% predictive routing accuracy and up to 20x faster network performance with Google Cloud Platform

Teridion Using machine learning to build high-speed virtual networks

Teridion, an Israel-based provider of an Internet Overlay Network for SaaS companies, speeds and stabilizes application traffic worldwide. After signing large customers and beginning to route hundreds of terabytes a day, the company needed a more scalable, automated infrastructure to eliminate geographic performance borders and handle massive streaming data.

Teridion migrated its Teridion Management System to Google Cloud—leveraging Compute Engine, BigQuery, and Cloud Machine Learning Engine—to move from reactive to predictive routing. The solution enabled real-time analytics and ML models that predict optimal routes with over 99% accuracy, accelerate traffic up to 20× and load data much faster, maintain customer availability during major outages, and scale to hundreds of terabytes per day; Teridion is now transitioning to a Kubernetes-based CI/CD architecture for further automation.

Teridion

Elad Rave

Cofounder and Chief Executive Officer