Case Study: Heureka Group saves 30% on cloud costs with CAST AI

Heureka Group reduces overprovisioning and saves 30% on the cloud

Heureka Group, Central Europe’s leading price comparison platform and online shopping advisor, used Google Cloud’s native autoscaler on GKE but found it often overprovisioned nodes, leaving more than 40% headroom they never used. To better align capacity with demand and reduce unnecessary cloud spend, they looked for a more dynamic autoscaling solution.

CAST AI provided autoscaling at both the node and workload levels, along with bin packing and Spot-aware optimization, to provision only the resources needed while maintaining performance. With CAST AI, Heureka Group achieved 30% cloud cost savings across Dev and Pre-prod clusters, reduced node pool resources by about 33% in Dev and Pre-prod, and lowered compute costs in Dev by quickly bin packing Spot-friendly workloads and removing empty Spot nodes.

Heureka Group

Martin Petak

Infrastructure Team Lead