Case Study: Barracuda Networks achieves faster phishing detection and stronger email protection with Databricks

Barracuda Networks uses ML on Databricks Lakehouse to prevent email phishing attacks at scale

Barracuda Networks, a global leader in security and data protection, needed a faster, more scalable way to detect phishing and impersonation attacks across millions of mailboxes. The company was building machine learning models to analyze email content and sender behavior, but feature engineering and deployment were difficult because labeled data was spread across time and code had to be duplicated between research and production. Barracuda used the Databricks Data Intelligence Platform, including Databricks Feature Store and Managed MLflow.

With Databricks, Barracuda centralized features in a single store, kept them continuously updated, and packaged models so they could be deployed more simply and efficiently. This streamlined ML operations and improved detection quality, helping Barracuda block tens of thousands of malicious emails each day and protect millions of mailboxes across thousands of customers.

Barracuda Networks

Mohamed Afifi Ibrahim

Principal Machine Learning Engineer