Carolina helps firms across all industries to “operationalize” their SAS programs, accelerate performance, respond to changing markets, and slash deployment costs.
Carolina converts SAS models and programs to Java, enabling deployment in any environment, including web services and operational systems (such as rules engines, CRMs, and decision support applications). This allows firms to perform model scoring using the most recent data on a just-in-time basis. Carolina also runs SAS programs in Hadoop providing significant performance gains over native SAS code. Scalability enables SAS programs to process big data efficiently.
Carolina allows firms to bring their SAS programs and predictive models to market significantly faster, in the operational environment best suited to their business requirements and infrastructure. By eliminating the time lag between model creation and deployment, firms can extend the shelf life of their analytical IP and respond promptly to the latest market opportunities and operational risks. Rapid deployment of the latest models is especially critical for companies launching new product promotions or adapting to changing fraud patterns.
Carolina significantly reduces labor costs by automating error-prone manual SAS code conversion and deployment workflows. Built-in validation ensures identical results between original and converted code. As an industry-standard framework that runs on all platforms, Java is both IT-friendly and far easier to operationalize than SAS. Carolina-converted programs can be supported by IT staff without training in SAS, and SAS-trained analysts and data scientists don’t need to learn Java in order to deploy their models at scale.