The ease with which an infrastructure-as-a-service deployment can be scaled makes the public cloud a prime candidate for hosting analytic applications that have to process massive amounts of information. Google Inc. last week moved to make it simpler for customers to take advantage of the benefits with the release of a managed machine learning service for developing data crunching algorithms.
The addition is meant to level the playing field against rivals Amazon Inc. and Microsoft Corp., which have provided similar functionality in their public clouds for quite some time now. Google hopes to stand out by enabling analysts to cleanse and filter the information they’re interested in processing using its Dataflow service before running it through their machine learning models. The functionality removes the need to utilize a third party solution for the task, which lowers software expenses while avoiding the hassle of constantly shuffling records in and out of the search giant’s infrastructure.
Oracle Corp. is likewise using the tight integration between its services as a selling point for its public cloud, which received a significant update against the backdrop of Google’s entry into the machine learning space. The enterprise technology powerhouse on Wednesday added several dozen new offerings to the platform including a data management suite aimed at simplifying the chore of operating off-premise analytics workloads. The arguably main highlight of the bundle is a discovery engine designed to let users find and import any useful information that might be stored on their organizations’ on-premise infrastructure.
Its broad feature set is enabling Oracle’s public cloud to win over a lot of new customers even as the vendor’s more traditional businesses, particularly its core database division, are seeing their revenue eroded by smaller players. Among them is Citus Data Inc., which last week launched a new version of its PostgreSQL distribution that is touted as much more efficient and reliable than the previous release. Its edge is credited in large part to a new automatic error correction function that is able to quickly perform troubleshooting if one of the nodes in a deployment fails or its data protection mechanisms encounter difficulties.