IBM’s advanced analytics portfolio contains its descriptive analytics, predictive analytics and prescriptive analytics products (everything from Cognos to SPSS to optimization and decision management). Within this portfolio is IBM’s decision optimization portfolio that contains the solver technology embedded in many IBM and external products.
IBM’s advanced analytics portfolio contains its descriptive analytics, predictive analytics and prescriptive analytics products (everything from Cognos to SPSS to optimization and decision management). Within this portfolio is IBM’s decision optimization portfolio that contains the solver technology embedded in many IBM and external products. This demand for decision optimization involves enterprise Operations Research (OR) teams, embedded solvers in software products and a series of business domains in which optimization plays a major role, such as scheduling, inventory management, etc.
Optimization has been a somewhat minor market compared with the broader analytics market, but as the analytics market moves to “prescriptive” IBM sees growing demand for decision optimization and the opportunity for mainstreaming of this technology. To take advantage of this, however, IBM believes that decision optimization needs pervasive ease of use. This means making it accessible to business people so it can be used in LOB applications AND making it work for data scientists who are not OR experts. IBM’s strategy for decision optimization has three steps:
- Self-service optimization where models can be dropped into an IBM cloud environment (announced recently) for rapid results and easy integration into business applications.
- End-to-end “model and solve” with advanced visualization, improved collaboration, model design UI, etc.
- Self-sufficient analytics where this is all integrated with other kinds of analytics, but non-technical users can use a marketplace to buy and sell services and solutions.
IBM feels that no one is really doing optimization right – too little integration, especially with analytics, too focused on OR experts, too big a gap between packaged applications and complete do-it-yourself optimization.
IBM is therefore offering:
- Its traditional optimization engine (IBM CPLEX) with an OR-centric IDE
- An application development platform for building optimization solutions, including tools to help you build scenario management, visualization etc.
- The newly announced cloud-based Optimization as a Service
- Advanced tools for Stochastic Optimization
The new cloud service offers drag-and-drop solve of optimization models developed offline in the cloud, an API to allow this to be integrated with applications in the cloud, and a community for access to experts and solutions. The service is free to try (limited time/number of models). All the results are displayed directly as text in the UI or returned as a JSON object from the API so it can be integrated with anything. There are various pricing tiers for the new service, a limited free trial, an hour-based on-demand price point, a subscription plan for defined numbers of hours and an enterprise reserved server.
IBM is a vendor in our Decision Management Systems Platform Technology Report. More on Decision Optimization Center can be found here and on CPLEX Optimization Studio here while you can sign up for the cloud trial here.
