Case Study: Acquia CDP achieves enterprise-scale, cloud-portable ML and automated cluster management with Qubole

AgilOne Machine Learning At Enterprise Scale

AgilOne is an enterprise customer data platform that runs complex supervised and unsupervised machine‑learning models for more than 150 brands (including Lululemon, Travelzoo and Tumi), processing vast, heterogeneous data and delivering millions of personalized predictions per client each day. Its challenge was to scale model training and daily scoring across AWS and GCP while eliminating prototyping and provisioning bottlenecks, simplifying cluster management, and providing reliable orchestration and automation for dozens of production models.

AgilOne partnered with Qubole to run optimized Spark, Airflow, and Zeppelin notebooks on a cloud‑agnostic platform with workload‑aware autoscaling, managed cluster lifecycle, and comprehensive APIs. The result was self‑service, automated provisioning and orchestration, faster prototyping and model deployment, consistent cross‑cloud operations, reduced reliance on operations teams, and the capacity to support near‑real‑time, large‑scale ML workloads (supporting the platform’s close to one billion daily predictions) with lower costs and strong enterprise support.

Acquia CDP

Gangadhar Konduri

Chief Product Officer