Case Study: Outbrain achieves around 40% infrastructure cost savings and scalable on-demand research clusters with Google Cloud Platform (Cloud Dataproc)

Outbrain Serving up some of the web’s best content with Cloud Dataproc

Outbrain, a leading content-discovery platform that makes roughly 275 billion recommendations a month, needed to modernize an aging research cluster that held about 6 petabytes of data. The company required a fast, non-disruptive migration that would work with its on-premises production systems, provide elastic capacity for research workloads, and give clearer cost visibility — all on a tight four‑month timeline.

Teaming with DoiT International, Outbrain built a hybrid research cluster on Google Cloud using Cloud Dataproc, Cloud Storage, autoscaling and preemptible nodes, migrating data via physical network links after a rapid POC. The move enabled seamless hybrid operations, faster project spin‑up, and clearer cost controls, cutting research infrastructure costs by about 40% while improving efficiency and attracting new engineering talent.

Outbrain

Orit Yaron

Vice President Cloud Platform