Cloud data protection company Druva today debuted a service that provides centralized management and data protection across public, private, and hybrid clouds. It also uses machine learning to make backups and recovery faster and more efficient.

The Druva Cloud Platform — the company calls it data management-as-a-service — combines the capabilities of its two earlier products including data protection, deduplication, governance, visibility, and analytics.

“Organizations have all these different data locales, and they’re moving workloads around based on what’s most cost effective and what delivers the best return for their business,” said Dave Packer, VP of product and alliance marketing at Druva. “In Druva Cloud Platform, we’re bringing together all these sources under a single control plane.”

The service backs up data stored in on-premises data centers and endpoints including laptops and mobile devices. It also supports Amazon Web Services (AWS) and Microsoft Azure public clouds.

“Would we add on more cloud services? Sure,” said Packer. “It’s just a matter of when those business cases come along and match up to the technology stack that does exist.”

The new platform includes a storage engine that uses machine learning to “auto-tune” data protection policies. “In this release, we’re starting to build in predictive backup, using machine learning to optimize the data transfer,” Packer said.

Not all data is the same age, or has the same level of activity or importance. Machine learning allows the storage engine to “auto-tune the policy to make it more efficient in each pass,” he explained. “You hear about machine learning and auto-driving cars, but then there’s things like backup recovery where it makes a lot of sense because there’s a lot of information to work with that’s predictable.”

The storage engine also provides a unified data fabric across platform services. This breaks down stored silos of information and allows companies to search, delete, recover, and monitor their full data set.