Analytics & Optimization Can rules help with SOA before “Level 5”?

Oct212005

In this blog entry - From ESB to ESP: action-oriented SOA - Joe McKendrick - discusses the SOA Maturity Model being promoted by some SOA vendors. The first time this 5 level model mentions rules is in level 5Level 5: The SOA information systems becomes the "enterprise nervous system" and takes action automatically according to events occurring at the business level according to rules optimizing business goals. So does this mean that business rules is a technology suitable only for the last level of maturity? I don't believe so. When getting started at level 1 business rules can be a huge boon. Not only are decision services an obvious first service - the desire to decide consistently across channels and across systems is often a key driver for SOA - but they also represent a great way to take part of an existing system and expose it as a service. I wrote about that approach here. When an organization reaches level 3 and starts to try and manage responsiveness then the use of business rules will start to show a strong return. Business rules are ideal...

2 Comments

Analytics & Optimization Robust Web Apps with rules-driven conversations

Oct202005

This article in ComputerWorld - No Pause to Refresh: More Robust Web Apps - Heather Havenstein talks about the new world of richer web applications. She says:The new tools essentially eliminate the need to refresh a Web page every time a user enters or receives new data... Companies are using the tools for e-commerce and call center applications that require complex interactions with users. So why do these new, complex interactions require a different kind of development approach. Well:Traditional HTML-based applications are built as static, page-based applications that require an HTTP request to a server every time data needs to be refreshed. However, applications on the Web often require heavy interaction with users. For example, a customer service application could require a call center representative to enter information from a customer while also accessing a technician's service schedule. These types of applications can't work efficiently with the latency that comes with constant page refreshes.What strikes me as interesting about this situation is the extent...

Leave a comment

Analytics & Optimization Using business rules to renovate old applications

Oct202005

In this article - Tough Decisions for Old Apps by Stephen Swoyer - there's a great summary of the problems of integrating legacy applications into an SOA:Web enabling or exposing mainframe applications to new and different constituencies—including non-mainframe applications—is supposed to be easy, provided organizations have prepared by separating their app’s business logic from its presentation logic. Unfortunately, some industry watchers claim, that's not the case for many mainframe apps, forcing enterprise service-enablement initiatives to negotiate some fairly difficult roadblocks.This is indeed a common problem with legacy systems. One example of how to address this can be found in this great case study from the California DMV - http://www.edmblog.com/weblog/2005/08/license_fee_cal.html. The basic process they followed can be used by anyone facing the issue of how to expose the core logic of a mainframe system as a service: Identify the services that are hidden in the legacy system Figure out the rules implemented in these services - perhaps by...

Leave a comment

Analytics & Optimization Customer Segmentation to increase profits

Oct182005

Customer segmentation is often the first kind of analytics attempted by a company. It is not, however, only for "newbies" - even the most sophisticated companies can use better segmentation to drive better results. Take Barclaycard, one of Europe's largest credit card companies and a very sophisticated company when it comes to analytics. They had seen consistently good results from their account management strategies, with “champion” strategies performing solidly. How then to do better? Their approach was to work on developing a better customer segmentation approach and then using more finely defined customer segments to better target the customers in each segment. They used a CART-based tool called Model Builder for Decision Trees from Fair Isaac , a tool that uses historical data to rapidly create new decision trees to segment customers using multiple performance dimensions. This ability to segment first, say, on retention and then on profitability and then on likelyhood to respond (rather than having to use a single perofrmance dimension...

Leave a comment

Analytics & Optimization Rethinking development with SOA and business rules

Oct182005

Saw an interesting blog entry today - Rethinking development - which talked about a report from the Burton Group. A couple of comments in particular caught my eye:Choose development tools and infrastructure products that support SOA design practices. Well that's pretty straightforward. Business rules products like Blaze Advisor with its focus on deploying rules services or decision services clearly meet this criteria. The blog entry goes on to say:[W]ith slow and steady commitment, SOA will decrease duplication of development work, simplify maintenance of service code, and improve overall consistency of business processes." The objective is "business flexibility and adaptability."This seems like a plug for business rules! Business rules management systems can reduce duplication by: Centrally managing business rules in a repository. Simplify maintenance by making it possible for business users to maintain their own business rules using a controlled but flexible interface. Make processes more consistent and simpler by automating previously manual...

Leave a comment

Analytics & Optimization Is a Services Architecture Threatening?

Oct132005

An interesting CIO insight article caught my eye today - Services Software Architecture: Efficient, but Threatening? This article suggested some good questions to ask about SOA and raised a number of great points about the potential for an SOA to be threatening. Let's take the questions first: Are our enterprise architecture efforts providing business users with the business flexibility they require? What is our plan for reducing maintenance costs and speeding up development times? Both of these questions could, and should, be answered by pointing to the development of business rules-based services. Such services not only have all the advantages of a service in an SOA, they also allow: Business users to maintain the rules in these services, giving them flexibility and agility Reduce time to develop complex decisioning services by using a technology designed specifically for that issue while taking advantage of the pluggability of SOA Dramatically reduce maintenance by eliminating the role of IT in changing key business rules in those services (see Gartner,...

Leave a comment

Analytics & Optimization Real-time or Right-Time

Oct122005

This article on real-time v right-time by Jim Ericson made me think about how this topic plays out in an EDM context. The key issue it seems to me is to establish which moves towards real-time offer real business value and which do not. I thought I would start with an old example of improving the time it takes to do something and how only certain improvements make a difference. Let's say I am a bank and it takes my competitor 5 days to make a loan decision (this is an old story, no-one takes 5 days anymore). If I can move my loan decision closer to real-time I should gain a competitive edge, right? Well, perhaps not. If I reduce my time to decide to say three days my customer still has to leave the branch and come back some days later. They probably won't see a gain of a couple of days as critical (unless they have some deadline). What if I can reduce it to one day - just have them come back tomorrow? Well, perhaps.What if I can reduce it to a few hours - come back later today? Still probably not that compelling. But what if I can reduce it to say 10 minutes...

Leave a comment

Analytics & Optimization Real-time, risk-based underwriting

Oct122005

When Unitrin's Kemper Auto and Home Group rolled out real-time, risk-based underwriting the benefits included improved financial performance, consistent decisions across channels and greater operational efficiency. Kemper found that automating this decision worked smoothly “behind the scenes” to enable consistent, real-time decisions and delivered the efficiency they were looking for. Often, the agent can write the policy while the customer is waiting in his or her office. Not only did this improve customer service it led to increased business with customers signing up more readily thanks to the immediacy of the decision. A year after implementation, Kemper’s combined ratio dropped by eight points. Along with other technology improvement initiatives, automating underwriting in this way is credited with helping Kemper become a more efficient underwriting organization. The unit also met itsgoals for reducing underwriting losses. Kemper continues to see improvements. Other gains included efficiency and better use of resources. The old system required a hefty...

Leave a comment

Analytics & Optimization Autonomic Computing – are we there yet?

Oct122005

I saw this article on Autonomic Computing in Information Week - In HAL's footsteps - and it made me think of how business rules are being used to build diagnostic and often self-diagnostic systems. The article talked alot about the need for vendors to integrate hardware, provide common standards etc. It seems to me though that once all the infrastructure is in place, there will still need to be a "smart" layer that can take decisions operationally. This layer would exhibit all the features of an Enterprise Decision Management solution: It would use business rules to record procedures, best practices, rules of thumb from experts as to how to respond to particular failures, how to interpret readings, how to select new routings around failed equipment. It would use predictive analytics to turn historical log data into executable predictive models. These might take temperature data and use it to predict the likelihod of failure of a piece of hardware or traffic data to predict a bottleneck. Other models might use Neural Network technology to...

Leave a comment