Viscovery Integration and Deployment

Easy model creation, flexible model application

Integration

Many applications require models to be integrated into an operational environment for applying the models record-by-record in real-time. In the operational mode, Viscovery is invoked through a APIs which provide the corresponding parameters of single records to the Viscovery One(2)One Engine and return the predictive results to the operational system.

The Viscovery One(2)One Engine can be integrated into existing application software to seamlessly deliver results to other programs that require them. Depending on the available hardware and the complexity of the model, this enables typical performances between 10,000 and 100,000 decision per second.

Many Viscovery clients have been surprised with the ease and flexibility with which Viscovery solutions have been integrated into their business processes. In many cases, full integration of Viscovery models has been completed within half a day. Take advantage of the usability and performance of Viscovery solutions to enrich your business processes!

Learn more about Viscovery Integration and Deployment

Viscovery OEM offering

Many enterprises in various industries can substantially widen their product and service offering by including data mining capabilities. By taking advantage of Viscovery’s integration flexibility and expertise in predictive analytics applications, they can expand their portfolio and increase market share.

Diagram

The lean, robust architecture of Viscovery allows easy and flexible integration of Viscovery functionalities into third-party products. OEM partnerships with Viscovery have demonstrated that existing software is efficiently enriched with data mining capabilities and that industry-specific verticals are developed through the combination of Viscovery with special-purpose products.

Leverage your product portfolio with data mining functionalities and contact us to learn more about Viscovery OEM options.

Viscovery key benefits in data mining deployment

  • Automated model creation and application, using reusable workflows as process templates
  • Lean, efficient architecture that facilitates speedy integration
  • Flexible model integration for quick return on investment
  • Real-time online integration of predictive models into multiple operational systems
  • High performance of up to 100,000 decisions/second

Automation capabilities of the Viscovery system for the deployment of data mining

Generally the model creation and the model application phases are distinguished. Initially, a Viscovery model is created interactively. Once a model has been generated, it can be applied using application workflows.

All completed workflows can be re-used as templates to modify, improve or repeat creation or application processes with the same data or with new data.

Workflow variations can be interactively performed again manually or using the “process automatically” mode available from Viscovery Profiler and Viscovery Predictor. In addition, the creation of data marts and models as well as model application can be fully automated using Viscovery Scheduler according to user-defined schedules.

Viscovery real-time integration examples

The Viscovery data mining suite and the Viscovery One(2)One Engine provide the basis for an efficient and competitive deployment of Viscovery for real-time applications, such as the following:

  • Web analytics
  • Call center applications
  • Real-time recommendation systems
  • Automated update of customer segments in CRM systems and customer data marts
  • On-line monitoring of industrial processes

back to top