Live from webMethods - Fabric 7 - The New Face of BPM - James Taylor's Decision Management

« Live from webMethods - Introducing webMethods Fabric 7 and What's Next for webMethods | Main | Live from webMethods - Driving Business Transformation Through Process Improvement »

November 07, 2006
Live from webMethods - Fabric 7 - The New Face of BPM

I am attending webMethods Integration World this week (ebizQ is a media sponsor) and blogging live from the sessions. Next up for me was Susan Ganeshan of webMethods talking about the Business Process Management capabilities of Fabric 7.

Processes have evolved over time with lots of patchwork systems with data and process steps moving around between systems creating a "Patchwork Process". Lots of different kinds of processes with wide variation between them in terms of what kind of data is involved, people to people v system to system etc. However, almost always has both people and systems interacting in processes. Susan quoted Tom Dwyer of Yankee group as emphasizing that the benefits of Business Process Management, Business Rules Management, Business Activity Monitoring and SOA can be multiplied if they are combined effectively. She also talked about Gartner's Business Process Maturity Model and how it moves from very straightforward to highly adaptive BPM.

Susan discussed some key themes and features for Fabric 7 and here is my dump of the ones that caught my eye:

Make it easy to identify and fix problems using integration and BAM facilities
Allows management and design of processes by combining portal, process and modeling tools into a single BPM environment that supports BPMN and BPEL and allows the definition of KPIs for measurement once the process is running.
Fabric 7 supports documenting the process, swim-lane by swim-lane, so that less technical users can see how their piece would work.
The BPM environment is designed to support both system to system and people to system interactions
Reuse was important to customers so offer integration with source code control system for asset management
Metadata is captured as process is designed and this is stored in the new metadata repository with query and searches to help you find reusable components and then allow drag and drop of these assets into new processes.
Codeless development of pages for managing data to replace very labor-intensive design tasks when trying to capture or edit data being passed around the process. The forms have some ability to manage data validation and retrieve data etc. Hopefully there is also the option to use more complex validation (based on Blaze Advisor) but I will have to check, although it sounds like you can.
Workflow includes routing rules based on roles and both dynamic and static rules. In addition these routing rules can leverage more complex rules running in the rules engine (e.g. "route to person X if this rule service says they are a gold customer and to person Y if they are not). The worklist management comes out of the box and supports Outlook, which sounds cool.
The business rules engine (Blaze Advisor) runs in the integration server alongside the process engine. Susan identified two main uses - complex processes involving complex decisioning rules and need to change the rules quickly. While this involves externalized rules they are still nicely integrated into the process design pallette. I think there are some more options for embedding decision services into an SOA but these are key reasons. Taken with the BPMS features this should deliver real business agility.
Fabric has some good flow and transformation capabilities already and reuses this in Fabric 7. The integration of the rule server into the integration server should mean this can also use rules in Blaze Advisor to handle very complex transformations.
Testing got an upgrade too. You can now insert test data into the process model and watch it flow through or set breakpoints and debug it in a more technical way all in the design environment.
Built on the BAM environment to do some roll-up monitoring and drill-down into the process to see which steps are causing problems and so on. Some nice integration of monitoring with events to help see what might have caused changes in monitored levels.
Metadata repository is used to do impact analysis as make changes and the Cerebra repository automatically managed the linkages between services so you can do impact analysis without having to do a bunch of manual updates as to what each service does. I am sure, though I don't know, that this is also going to work with the Blaze Advisor repository to bring rules into this environment.