Case Study: How Attentive revolutionized personalized marketing with Anyscale
Key results
The challenge
As Attentive evolved its AI initiative, it ran into the problem of scale: building more complex models required processing and training on substantially more data than its home-grown, self-managed Kubernetes setup could handle. The team could not consolidate its data into a single model because it could not process it all.
The solution
Attentive adopted Anyscale as its AI compute platform, using its hosted workspaces, managed compute, autoscaling, and spot support to unify multiple models into one and train on far larger data volumes.
“Before Anyscale, we didn't have the ability to consolidate our data into a single model because we simply couldn't process it all. With Anyscale, we were able to unify it into one model and reduce the cost by 99% while increasing the data volume for it by 12X.”
CSChristian StanoEngineering Manager, ML Platform, Attentive
The results, in context
By unifying its models on Anyscale, Attentive reduced cost by 99% while increasing the data volume for that model by 12x. The team also achieved a 5x reduction in training time and a 50x increase in the number of customers supported by its models, and onboarded engineers to their first Anyscale workspace in 3 days.