Table of contents for IBM Analytics Analyst Summit 2012
Deepak Advani came back on stage to give more details on Decision Management software from IBM. Obviously Decision Management 6 came out a while ago, bring rules and analytics together. With the new release there are several new features:
- Integration of CPLEX
The new release integrates the ILOG CPLEX Optimization engine
- Entity Analytics
The integration of new entity analytics capabilities into SPSS and thus Decision Management is a new capability
- New analytic techniques
Research has provided some new algorithms such as social network analysis and analytics for temporal causality
- Real-time implementation in InfoSphere Streams
A new deployment options allowing the decisions to be deployed in InfoSphere Streams
Native deployment to z/OS
Plus, of course, there is the fact that IBM is making this a strategic play. And this is reflected by the support from different groups in IBM such as the Big Data Platform folks and the WebSphere team.
For instance, Arvind Krishna who heads up the Information Management team at IBM, talked about the need to deliver value from all the information sources available to enterprises. Big Data means not just more data but different kinds of data and rapidly changing data. To deliver business value from this IBM sees Decision Management working against in parallel with the Big Data platform.
Marie Wieck, who runs WebSphere meanwhile, talked about the next generation of business rules – IBM WebSphere Operational Decision Management. She used a range of examples to show how decision management matters – applying thousands of rules in processing Visa transactions, scaling expertise for athlete management and more. Part of her focus is also on real-time decision management by linking event processing and decision-making rules together.
IBM has built up a lot of IP around Decision Management too. This has resulted in some “Signature Solutions” around Customer Next Best Action, Finance Performance Insight and Anti-Fraud, Waste and Abuse. In the example of CNBA for instance, the request comes in from a channel to a Decision Service. This pulls offers from a marketing system like Unica, applies predictive analytics and rules to figure out the best next action for this customer. This is then delivered back to the customer in the right channel. Clearly GBS, IBM’s consulting force, is also getting behind Decision Management pulling together the various elements of software, analytics, research and more.
Finally, Watson represents an interesting Decision Management angle for IBM.
The preventative maintenance scenario that Deepak mentioned was then used to give a demo. Some key points from the demo:
- Decisions being made automatically and being fed into a dashboard to show a supervisor what the system did
- Rules and predictions being applied together to make decisions about maintenance work required
- Rules apply the right engineer to fix the problem but only once a simulation shows that the resources will be effectively applied
- Optimization applied to see what the impact of additional engineers might be, what the value of this would be
- What if comparisons let you see which one will work best
- And then everything can get wrapped up and deployed so that all the manager needs to see is the result in their dashboard.
- Decisions as a Service (Answers as a Service as IBM calls it), is coming (in pilots) while making Decision Management available on a SaaS offering, is available now
- Working with IT is a journey – needs to work through how analytics can help with each specific business problem, each specific decision.
- Can integrate analytics into rules-based approach or rules into analytic decisions – IBM WebSphere Operational Decision Management and IBM SPSS Analytical Decision Management are integrated.
- Real-time decision management requires real-time scoring and real-time application of business rules but offline development of both.