RapidMiner today announced a new Platform-as-a-Service offering for deploying and managing predictive analytic models in the cloud. RapidMiner consists of a Studio product for developing analytic models and a Server product that allows for collaboration, batch model execution and interactive dashboards etc. I have blogged about RapidMiner before and the product is in our Decision Management Systems Platform Technology Report.
The new product makes RapidMiner Server available as a Platform-as-a-Service offering. A cloud-based server can be enabled for a client quickly, allowing them to have a full fledged server, providing all the capabilities of the RapidMiner Server product (including any extensions that have been added), in the cloud. The RapidMiner Server PaaS is hosted on Amazon AWS and gives clients both a server that they don’t have to manage and one that is closer to their cloud data sources (Amazon RDS is an increasingly common location for data sources these days).
The RapidMiner Server allows any RapidMiner analytic modeling process definition to be converted to a web service. With the new PaaS offering these web services are immediately available as RESTful services running on Amazon AWS. These services can run in batch against many records in response to triggers or at specific times. These deployed processes can also be used to build web-based interactive dashboards using the RapidMiner dashboard capabilities. All the RapidMiner Server management tools are browser-based and this allows the server itself, users, privileges etc to be managed remotely.
RapidMiner is hoping to attract new users such as small organizations who would like a server but don’t want to engage in the IT tasks involved in managing a server. In addition many of their existing server customers are also interested either because their data sources are increasingly running in Amazon’s cloud or because they are making a general move to the cloud. Most users of the RapidMiner Server initially focus on using the server to improve their model building process – using the server to build the models and collaborating on and managing the models – but there is increasingly strong interest in making the models available in production too. Similarly, while most users are using the services to get information to display in a dashboard or a report, there is already a fair amount of activity in programmatic access to the services in more of a Decision Management context.
You can get more information on RapidMiner here.