SAS Open Model Manager

SAS Open Model Manager optimizes management of analytical models to solve a problem of "the last mile".

2019: Release of the SAS Open Model Manager application

On October 24, 2019 it became known that the SAS company announced release of the SAS Open Model Manager application which will help to unroll analytical models open source, and then to conduct monitoring of their work, to control efficiency and to timely reveal signs of obsolescence. The solution was created especially for community of software developers open source therefore for its start and work with it no other technologies of SAS are required.

Many organizations have problems with the so-called "last mile" in analytics when solution development is not completed by his output in operation. Partly it is connected with labor input of manual processes, partly – with unstable interaction between IT departments and business users. According to the research IDC, less than a half of the organizations can claim that they adequately use the analytical models, and only 14% say that results of work of their analysts and data scientists are involved completely. Thus, in the majority of the organizations the potential of analytics remains strongly underused.

The solution SAS Open Model Manager helps with a unified environment to adjust work of data scientists and IT specialists, adjusting processes of implementation and control of models open source, including popular Python and R. The built-in subsystem of the reporting allows users to estimate utilization properties of models, to compare them and to quickly make decisions concerning preserving or removal of old models and development of new.

The publication and categorization of models can pass in different operating environments as in batch mode, and in real time. The solution also simplifies accounting and management of models, helping users it is better to understand functionality and utilization properties of the implemented models in dynamics.

Access to the solution SAS Open Model Manager and its deployment will be performed using container technologies, including Docker and Kubernetes. It provides portability, ease and an opportunity to start the solution in a private or public cloud.

SAS also announced ModelOps, other component designed to solve a problem of "the last mile". This package offer allows the organizations to manage models and to scale them according to requirements and also to continuously conduct monitoring of their work, to reveal signs of decrease in efficiency and to quickly make corrections.