Case Study: CrowdStrike achieves rapid scalability and agile big-data ML processing with Amazon Web Services

CrowdStrike - Customer Case Study

CrowdStrike is a cloud‑native leader in endpoint protection whose Falcon platform uses machine learning and behavioral analysis to stop breaches. The company needed more agility and scalability to run Spark‑based ML workloads on hundreds of terabytes of event data, a flexible and cost‑effective way to store petabytes of Cassandra data, and higher availability so it could quickly rebuild or scale infrastructure without long maintenance windows.

CrowdStrike moved its processing and storage to AWS—using S3 for ingestion, EMR with Spark for big‑data processing, EC2 (C4) instances and EBS for Cassandra—and gained the ability to spin up clusters on demand, reduce operational overhead, and cut storage costs (EBS at roughly one‑third the cost of SSD instance storage). The change enabled rapid scaling to meet growth, supported high write loads and geographic redundancy across availability zones, eliminated disruptive downtime for infrastructure changes, and accelerated model validation and feature development.

CrowdStrike

Sven Krasser

Chief Scientist