Abstract: A device trains a machine learning model with historical productivity data and skills data to generate a trained machine learning model that determines allocations of tasks to workers. The device receives new task data identifying new tasks to allocate to the workers and performs natural language processing on the new task data to convert the new task data to processed new task data. The device receives, from sensors associated with the workers, real-time productivity data identifying productivity of the workers in completing current tasks assigned to the workers. The device processes the processed new task data and the real-time productivity data, with the trained machine learning model, to determine allocations of the new tasks to the workers, and causes the new tasks to be allocated to the workers by one or more devices and based on the determined allocations of the new tasks.

Abstract: A data analysis apparatus generates M (M is an integer of 3 or greater) groups each including data regarding a plurality of units from data where first KPIs and K (K is an integer of 2 or greater) explanatory variables are given by 1:1, generates a second KPI indicating the state of the group based on the values of a plurality of first KPIs included in the group, and selects a feature for the first KPIs based on a correlation analysis between the second KPI of each group and the feature of each group calculated based on the explanatory variables.

Abstract: Provided is a user interface for generating virtual models of configurable products. In an example, a method includes displaying the user interface including a workspace and a plurality of functions for generating models of configurable products, generating a virtual model of a configurable product based on input received via the user interface, the virtual model comprising attributes selected by the user input which are capable of being dynamically configured to generate a price quote for the configurable product and a relationship between the attributes, and in response to the virtual model being generated, generating syntax that corresponds to the virtual model and storing the syntax in an executable file. The example embodiments provide a user the ability to quickly and easily generate CPQ objects that can be executed using CPQ software without requiring the user to learn CPQ syntax.

Abstract: Embodiments of an architecture for large scale key factor optimization and management are disclosed. The architecture may include an optimization and management system that uses existing data to characterize items, group them based on user selected similarity criteria and/or user-selected objectives for the performance metrics, further allowing the user to define target objectives for each group and through forecasting of the impact of the key factor value changes, select the adjusted key factor values that better achieve the selected objectives.

Abstract: According to an embodiment of the present invention, a system dynamically optimizes a service workflow including one or more software services. Initially, the system receives a request to optimize a service workflow, wherein the request includes one or more service requirements and one or more performance objectives related to the one or more service requirements. The system validates service plan information associated with at least one software service plan offered by one or more service providers and generates a plurality of candidate workflows based on the service requirements and performance objectives. The system then determines an optimized workflow from the candidate workflows and executes the optimized workflow to perform operations of the software services. Embodiments of the present invention further include a method and computer program product for dynamically optimizing a service workflow including one or more software services in substantially the same manner as described above.