Case Study: Abnormal AI achieves 20% fewer successful email attacks and 40% lower infrastructure costs with Databricks

Stopping sophisticated ransomware in its tracks

Abnormal Security, a cybersecurity company focused on stopping sophisticated email attacks with behavioral AI that analyzes over 50,000 signals, faced rapid growth and increasing demand that their legacy Hadoop/EMR infrastructure couldn't scale to meet. They needed faster analytics, more reliable data pipelines, and the ability to iterate on machine learning models to detect targeted phishing and ransomware at enterprise scale.

By migrating to the Databricks Lakehouse Platform (Delta Lake, Databricks SQL) and ingesting streaming signals for near-real-time analysis, Abnormal removed heavy infrastructure overhead, enabled collaborative data work, and deployed high-performance ML models that process thousands of emails per second. The move cut successful email attacks by 20%, reduced infrastructure costs by 40%, and boosted data team productivity by over 30%.

Abnormal AI

Sanny Liao

Head of Data Science