SAS® Model Manager Speeds Value of Predictive Analytics

Today's organizations rely on analytical models to make strategic and tactical decisions, such as estimating risk exposure, analyzing business strategies and predicting customer profitability. Models play an increasingly important role in business processes, and erroneous model output or incorrect interpretation of model results can cause million dollar mistakes. 

But a new SAS offering, SAS Model Manager (shipping in November 2006), will help deploy accurate analytic results to decision makers worldwide.

Lifecycle analytics
As more analytical models are implemented into operational systems, model decay becomes a challenge. Retaining poorly performing models in production status is bad for business. The SAS enterprise model management framework supports best practices for organizing and tracking the tasks of model development:

• Model verification and testing.
• Comparative model performance benchmarking.
• Publishing and sharing model performance data and metrics over established reporting channels.
• Retirement of the model when it's pulled from production status.

SAS calls this "lifecycle analytics."

"This tool offers a level of control and standardization for our modeling process that we haven't seen anywhere else," says Robin Stewart, Senior Manager Customer Acquisition and Development for BMO Bank of Montreal Cardholder Services. "The amount of time spent managing the life cycle of our models would be greatly reduced by using SAS Model Manager, saving us valuable time and reducing our total cost of ownership."

SAS Model Manager in action
At Swisscom, a leading telecommunications company in Switzerland, SAS Model Manager centrally organizes, classifies and assesses models for them and thereby reduces complexity. With SAS Model Manager, it's easier to determine which model is best used in which situation and to ensure correct results.

Effective analytical models are high-value assets and must be well-managed for optimum performance. This new SAS offering is the best defense against such "model risk" with its robust validation and detailed systems of checks and balances that continually verify the accuracy and usefulness of the analytical models. SAS Model Manager enables the transfer of analytics from the environment where they were originally created to the production operational environment with a consistent, repeatable process.

"With this latest offering, SAS is well-positioned to address the whole process of model creation, management and deployment for optimal model performance," says Dan Vesset, Research Director for IDC.

Bridging the gap
"As we enter a new age of operational analytics and an increasing regulatory environment, leading organizations understand the requirement to treat the management of analytics as a life cycle," says Keith Collins, CTO at SAS. "Whether your portfolio is a couple of models or a couple hundred models, there is a real opportunity to streamline the manual implementation process. SAS Enterprise Model Management elegantly addresses that need, bridging the gap that has existed between the model development and the model scoring environment. It's the only software on the market with a patented, secure model repository and rich metadata structure to enable the deployment of analytics."

SAS Model Manager is integrated with award-winning SAS Enterprise Miner™, a rich suite of data mining tools that empowers users to explore and exploit corporate data for strategic business advantage. SAS Enterprise Model Management is part of the SAS Enterprise Intelligence Platform, which optimally integrates individual technology components within an existing IT infrastructure into a single, unified system. The result is an information flow that transcends organizational silos, diverse computing platforms and niche tools – and delivers new insights that drive value for your organization.

SAS Model Manager will streamline creating, managing and deploying analytical models.