Skip to main content
Live from Brainstorm (almost): Business Process Management, SOA, and Composite Applications

Ken Vollmer of Forrester Research gave a nice presentation on Business Process Management, SOA, and Composite Applications.  I particularly liked both the phrase "Digital Business Architecture" to describe these general trends and his comment that we are finally able to treat processes as "business metadata". Clearly I think EDM is a core component of the first phrase and that a business rules management system let's you treat business logic, likewise, as business metadata.

Some specifics takeaways:

  • Ken outlined steps of Define, Implement, Execute, Monitor, Optimize and back to Define as the loop you need to create. Business rules play a key role in allowing you to apply your optimizations to your deployed processes.
  • He talked a fair bit about complex event processing (CEP), something for which I think rules have great potential.
  • He discussed how self-healing processes can be developed that use business rules, analytics and CEP to impact running processes without the need to wait for someone to review a dashboard.
  • He noted that there were two broad (human- and integration-centric) and four detailed categories of business process management suites and that companies might find good reason for using multiple. It seemed to me that the idea of adopting "one engine to rule them all" (apologies to Tolkien and a GIGA analyst) would be key - if you used a common rule engine for your core business logic, it would mitigate many of the risks inherent in using two BPMS.
  • The ability to configure services was noted as key to reducing costs. Clearly services built with business rules management systems are much easier to configure than those written with code.
  • Ken gave a nice case study of a mortgage origination system as composite app. Me, of course, focused on the fact that it still had a person making the decision (where rules and analytics could have automated a huge percentage) and still went and got external data (which must be paid for) before running any kind of decision logic to see if it would actually matter - why pay for a report on the house if you know you won't underwrite the applicant?

Lots of good stuff in Ken's presentation, just a sampling here of my key takeaways.

related posts