Named Amazon Machine Learning, the fully managed service will enable developers to create predictive models based on historical data, and apply it to a variety of situations, including detecting problematic transactions, customer turnover, and improving customer support.
According to Amazon, Amazon Machine Learning’s APIs are based on existing technologies already deployed by developers on AWS. The new service will be designed for ease of use with wizards and guides for deploying analytic models.
“We use Amazon Machine Learning to analyze customer feedback on packaging and create predictions to identify products that are suited for our Frustration Free and eCommerce ready packaging standards,” said Kara Hurst, director of Amazon global sustainability in a statement. “Amazon Machine Learning has really helped us improve our ability to identify products with packaging that is wasteful and frustrating for our customers.
The service would be integrated with data already stored in Amazon’s cloud, and integrates with Simple Storage Service (S3), Redshift and Relational Database Service (RDS). According to Amazon, the service has no setup cost, and is pay-as-you-go.
Meanwhile, IT providers MicroStrategy, Software AG, TIBCO, and Onshape are joining a growing list of partners who have implemented AWS as their core infrastructure platform for their SaaS offerings.
“Software AG chose to go ‘all-in’ on AWS because they offer outstanding operational efficiency and the widest geographical footprint, enabling Software AG to focus on our core offering – building a highly differentiated, value-added solution for our customers across the globe,” said Wolfram Jost, CTO at Software AG in a statement.
Lastly, AWS announced its Marketplace for Desktop Apps.
According to the company, it’s a new category under the marketplace where users can find desktop applications for Amazon WorkSpaces. More than 100 applications in 11 categories, AWS said.
The cloud provider is even implementing a new service that bundles applications and delivers them to WorkSpaces as though they were natively installed, but are administered by IT.