Cambridge-based predictive analytics company RapidMiner announced on Tuesday that it has acquired Radoop, an analytics company that optimizes the big data platform known as Hadoop. Prior to the move, RapidMiner already had an existing partnership with the company, incorporating Radoop’s enterprise-geared big data analytics products into RapidMiner’s suite of predictive analytics offerings. With the acquisition, more of Radoop’s Hadoop tools can be accessed by RapidMiner’s customers, building a larger presence in the Hadoop ecosystem for RapidMiner. The acquisition also brings partnerships with Hadoop platforms Cloudera and Hortonworks, and adds 20 new clients to RapidMiner’s customer base.
RapidMiner, a big data analytics startup, purchased former partner Radoop, giving it Hadoop processing and access to the Radoop team’s Hadoop expertise. They did not share the purchase price. RapidMiner enables customers to create a data-analytics workflow and grab data from a variety of sources without scripting using built-in connectors, then process that data to see patterns that might require action. It then offers a more detailed view to see possible solutions to issues that arise. For example, if it’s analyzing data from a CRM tool, says RapidMiner CEO Ingo Mierswa, it might learn from the data that the customer can expect a 30 percent churn rate in the next year, but it doesn’t stop there. He says it will then start analyzing each customer and predicting the likelihood it will churn to provide predictive actions at a detailed level. After RapidMiner completes the analysis, customers can send the results to a business intelligence package, or deliver results as part of the workflow design. Until purchasing Radoop, RapidMiner processed the data in-memory or executed directly in the database. With Radoop, customers now have the option of processing on a Hadoop cluster, so instead of being limited to a single machine, they can add multiple nodes, giving customers the ability to process much larger data sets than they could before (without using a partner product like Radoop).
