Still catching up on some webinars that I missed, here’s one from June 30th about BPM and Business Rule Management Systems (BRMS). It’s hosted by Peter Carter of ILOG, and although he’s not the most dynamic speaker that I have heard, there’s some interesting content on the slides for those of you who might be working through the relationship between BPM and business rules:
- BPMS focus on how people and systems interact
- BRMS focus on how decisions are made
His point is that BPM can get you to a certain point in process automation, that is, where the only human intervention is for decision-making; BRMS can take you further to automated decision-making so that the only human intervention is for exception-handling. Of course, you can embed rules within a process map in any of the BPMS that I’ve ever worked with, but the rules are typically not explicit and may have to be replicated at a number of points in the process map, depending on the capabilities of the BPMS: most of the systems are really not built for explicit rules management. And, although all the BPMS vendors claim that it’s easy to have a business analyst change the process map at any time, in reality I’ve never seen an organization that would allow that to happen without (justifiable) layers of technical review and testing to ensure that the new process logic doesn’t break something. As I mentioned in this post, you’re pretty unlikely to find someone within the business part of an organization that is sufficiently trained/experienced in systems thinking to be able to make what are essentially programming changes, even if they are using a pretty graphical interface.
Business rules typically interact with a BPMS at a specific point in process, usually to generate a value and/or determine the next action to take based on the current parameters (e.g., pricing, eligibility determination, scoring, case assignment, data validation). That makes it a lot easier, and a lot less dangerous, to allow a business analyst to modify business rules without an undue amount of IT involvement: since the process map is not changed, there is no danger of work becoming orphaned or misdirected through a business rule change as long as the process map was properly designed in the first place.
This is why all industrial-strength BPMS are either OEM’ing a business rules engine into their product these days, or integrating tightly with one: you can’t swing an analyst’s report these days without hitting some new BPMS/BRMS alliance. ILOG’s certainly a driving force here, with alliances with Siebel, FileNet, Oracle, IBM, BEA, Fuego and a host of other major players. There’s one major exception to this alliance trend: Pegasystems’ BPM Suite is natively built on a rules engine, which has given them a leg up on the process+rules craze.
Going back to Mr. Carter’s presentation for a moment, I also like the way that he defines business rules from both a business and IT perspective:
- From the business perspective, business rules are “formal statements of business policy that define or constrain some aspect of the business”
- From the IT perspective, business rules are “business logic to be implemented in one or more business applications”
Too often the definition is made from only one of these two perspectives and, like Mars and Venus, someone doesn’t fully understand what was said.
If you want your business rules integrated with your processes, you’re going to have to figure out how your BPM tool and rules engine will integrate, then pull your rules out of the paper procedure manuals, mainframe applications, spreadsheets and all the other places where they live, and get them into that BRMS. Easier said than done, but rest assured that rules engines are here to stay in the world of BPM.
This may not be the best venue to raise this query. Nonetheless, I need to know about the manner in which use-cases would typically integrate with business rules and the way business rules are specified during application design. An answer would be much appreciated.