Your data is distributed across the enterprise, so should your analytics

The data in your business systems are distributed across the enterprise. Collaborative Analytics provides an architecture whereby the predictive analytics derived from these systems are microservices that are then integrated by the platform into enterprise-wide prediction network services. This modularization of services allows teams of data scientists to mine individual business systems independently, focusing on separate but related business goals in accordance to their individual domain expertise. The Collaborative Analytics platform provides the elastic fabric that integrates these microservices on demand to provide flexible solutions for evolving business goals, avoiding the expense of re-architecting the solution as the business scales. The advantages of this microservices architecture extend to the following cases:

• Evolving Business Systems: Organizations are continually motivated to improve internal processes and information systems to improve efficiency and stay competitive. Collaborative Analytics provides a valuable degree of abstraction for predictive analytics allowing piece-meal upgrading of possibly heterogeneous business systems using best-of-breed approaches without the pain of re-architecting the monolithic data model. The result is future-proofing your business systems as your business grows.
• Hybrid Cloud Infrastructure: While organizations are moving more of their IT infrastructure to the elastic resources afforded by public cloud services, they also recognize the need for data security and control provided by a private cloud. Collaborative analytics provides a platform that seamlessly integrates predictive analytics derived from data across clouds.
• Merger and Acquisition: Collaborative Analytics accelerates the use of information residing in new systems that are acquired through M&A activity, prior to the expense and time of integrating those systems, and help to prioritize what information would be most useful to integrate first.
• Legacy Systems: The model abstraction required by Collaborative Analytics allows integration without the need to understand the internal details, which means that it applies even to legacy rule-based models not learned from data. The only requirement is the evaluation of these models to understand their error-rate performance.

Collaborative Analytics is a form of Distributed Artificial Intelligence (AI) where intelligence derived from local models are organized and configured by autonomous agents in a bottom-up approach resulting in the discovery and emergence of prediction networks as the enterprise evolves

As a multi-agent system, the Collaborative Analytics platform evolves and scales as your business grows, providing:

• Continuous Learning and Integration: New sources of data and intelligence derived from that data can be incrementally deployed without interruption to on-line use of prediction networks. Plus, as local models are refined and improved, prediction networks for which the local model is a participant are automatically updated.
• Resiliency to Component and Network Faults: The learning of prediction networks bakes in optimal strategies to account for missing information from one or more agents during prediction. This robustness extends to an optimal treatment for missing data at one or more sources during prediction.