Carolina helps companies across a wide range of industries “operationalize” their SAS code to minimize costs, gain new insights, and respond faster to market demands. Learn how Carolina is used in the following markets:

Pharmaceutical manufacturers and biotechnology companies are leveraging growing volumes of data to drive drug discovery, reduce manufacturing costs and tailor innovative new therapies. While biostatisticians have long used SAS to design experiments and analyze clinical trial data, a number of new use cases have emerged that challenge the capabilities of traditional statistical tools.

Pharmaceutical researchers are using predictive models and advanced analytics to:

  • analyze vast volumes of genomic data to uncover hidden patterns and correlations between genetic anomalies and diseases,
  • mine social media posts for early detection of epidemics or other public health threats, and
  • optimize costly and contamination-prone drug manufacturing processes.

Effectively leveraging the disparate sources and volumes of data to drive more informed decision making requires the ability to operationalize predictive models and derive analytical insights quickly. Carolina technology provides pharmaceutical manufacturers and biotechnology firms with a highly scalable and automated solution for rapidly deploying SAS-based models and advanced analytics.

CASE STUDIES:     Drug Discovery     pharmaceutical manufacturing

Statistical and actuarial analysis forms the foundation of insurance underwriting, and explains why insurers were among the first industries to embrace the potential of Big Data. Insurers leverage massive volumes of meteorological, public health, casualty and economic data to inform their decision making. Telematics data collected from on-board vehicle sensors are the latest innovation insurers use to create custom policy pricing and detect theft and fraud. Predictive and classification models built in SAS are an integral component of insurer business processes.

But SAS models are difficult and often impossible to scale, requiring time consuming and error-prone manual workflows to convert to more suitable software environments such as Java. This leaves insurers at a competitive disadvantage, as models built on the latest data can’t be deployed into operational underwriting and fraud detection systems on a timely basis. Carolina technology alleviates this critical problem, seamlessly bridging the gap between model creation and deployment, by automatically converting, validating and deploying SAS-based models from within the firm’s operational systems.

Case study

Financial Services
Credit card issuers, banks and financial services companies are facing unprecedented margin pressure from fee-sensitive consumers, a proliferation of competing loan and credit card offerings, and increasingly sophisticated on-line fraud. In order to reduce costs, differentiate their product offerings and combat fraud, financial services firms generate and maintain massive libraries of SAS-based predictive models.

But models, no matter how accurate, have an increasingly short shelf life. Lacking the ability to deploy their latest analytical insights quickly, firms lose competitive advantage and expose themselves to mounting financial and operational risk.

Additionally, the growing volumes of data used by predictive models to target consumers can significantly hamper the ability of traditional credit scoring platforms to produce timely results. Carolina technology helps financial services providers rapidly deploy SAS-based predictive models at Big Data scale in Hadoop and other massively parallel frameworks, while eliminating error-prone manual workflows.

Case Study

The unprecedented growth of global e-commerce companies has been driven in large part by their ability to capture and analyze massive volumes of transactional, financial and consumer data generated on their websites. In addition to recommendation engines that drive cross-sell opportunities, e-commerce companies also create and maintain thousands of predictive models to detect fraud, optimize product pricing and discounting, and detect shifting trends in consumer preferences to manage their supply chains.

Many e-commerce firms use the SAS platform for predictive modeling and advanced analytics. But SAS models and programs pose severe portability and scalability issues, requiring significant time and manual effort to convert and deploy into operationally suitable environments such as Java. This results in lost revenue, stale merchandise, and potential exposure to the latest online fraud tactics. Carolina technology helps firms alleviate these challenges by providing a fully automated and scalable solution for deploying SAS models and analytics from within web services and operational systems.