Posts Tagged ‘Case Management’

So here’s a BPM capabilities matrix from the not-too-distant past. I won’t say which two BPM platforms were evaluated for the below matrix, much less the client and the effort, but I’m sure many have seen this format before and it’s intended as just a/one starting, jumping off point.

As I work on this BPM/Case Management series in the coming weeks and months (yes, it will take that long) I hope to blow this out, draft multiple use cases for structured and unstructured process both and, ultimately, draft a white paper for the products examined that goes way deeper under the hood than anything Gorrester or Fartner have ever done.

I’m soliciting any and all input to this artifact and the use cases both so I make sure I properly address the platforms’ technologies for how they perform any given functionality, if that’s good or bad, and why. Feel free to comment here or ping me on Twitter with your input.

I have a LOT of resources, references that I’m going to go through, use for this plus, of course, running VMs for all the platforms that will be examined (interrogated? <g>). I want to make sure I get it as comprehensive and unbiased as possible.

Cheers, Pat

        Graphical User Interface
        Icons to represent Steps
        Connections between steps
        Drag & Drop
        Reusable Submaps
        Process and Property Inheritance
        Model Management
        Fine grained security
        Public & Private Inbox/Workbasket & Worklist
        Split and Merge steps
        Spawn new workflows
        Interfaces – Web Service, Human, Queue
        Import standard Modeling Languages – BPMN, XPDL
        Visio Interface
        Resource Allocation
        Work shift creation and assignment
        Intelligent metadata repository
        Intelligent Recommendations
        Rules Engine communication
        Dynamic workflow customizations
        Decision making
        Intelligent Routing
        Skill based routing
        Content based routing
        Document Management Functionality
        Load balancing work
        Event Driven
        Intelligent workflow activity traversals
        Integration with external systems
        Ad-hoc reporting
        Scheduled reporting

So what’s the moral of your story Mr. Peabody?

  • This is gonna be bigger than a breadbox.

What’s next?


Read Full Post »

64KWhen examining business and technical trends, particularly with today’s social media hyperbolic echo chambers’ constant blare, it is instructional to look at history and understand how we got to where we are.

In the case of IBM Case Manager (ICM) I first noted, began paying attention to it in late 2010, reaching out to IBM ECM Education and saying that as soon as there was a class for it I was interested. (Sidebar to the technologists: Always try to stay in the vanguard, bleeding edge pays. I’ll explain more later). So, in the fall of 2011 I trotted off and took IBM Advanced Case Manager (ACM) Installation concurrent with attending FileNet P8 5.0 Installation that particular week. I took the latter class with a then-client present in the room and will hit them up in the #SoC series later.

Back then it was Advanced Case Manager and dependent upon the vendor the ‘A’ stood for something else – ‘Advanced,’  (IBM), ‘Adaptive’ (ISIS Papryus, more on @maxjpucher later. Remember the history part on this series throughout) or even a different letter – Dynamic Case Management (Pega). We’ll come to each of those in turn. On the Big Blue side of things suffice to say that within the course of the past four, five years everything has come to have an “I” in front of it. Go figure.

Flash forward to current times and everybody pretty much generically just says “Case Management” and leaves it at that. 😉

ICM 5 InstallationTurning back to ICM c.2010/2011 the installation chronology for installing, configuring and deploying the apps (plural) was thus. For ICM it’s important to remember that in FileNet P8 5.0 Process Engine (one of the three app servers in the base “triad” of P8) was newly made a Java (not J2EE) app and that Content Engine and Workplace/XT were JEE apps on middleware, as was/is ICM.

ICM 5_2And, moving forward in time five years, PE has now been folded in with CE to be “CPE,” Content Process Engine, on middleware though the table structures in the underlying DB(s) are still there. More on that later as we dive down as that aspect’s importance, relevance is examined vis-a-vis the other platforms we’re going to look at. However, we still have three .EARs and one .WAR deployed out on WebSphere. By the way, the discourses on ICM are pretty much going to look at it from a “blue-washed” perspective so you can pretty much assume WebsFear throughout this series whenever I talk about ICM.

Now, though users and most people who aren’t on the bare metal probably don’t care that much about a bottom-up, systems architecture view on any of this I give it to you, point it out because whether they know, care, or acknowledge it or not a platform’s legacy and architecture very, very much determine how it does things the way it does. How the technology effects the business functionality, which things it does well, how and why. All of these platforms and their technology have idiosyncrasies and what I like to call “tips, tricks and traps.” All have been around long enough now to have an empirical body of knowledge and history of what constitutes best practices for that given platform and its underlying technology.

For this opening salvo I just want to point out that, in the case of ICM, it is an application (dare I say “framework?”) that sits on TOP of IBM Case Foundation (nee IBM FileNet P8) and needs ICF in order to do its thing. Other than a couple of kinks I’ll point out, the same is not true of the other platforms we’re going to examine, dive down into. Each has their pro’s and con’s.

So what’s the moral of your story Mr. Peabody?

  • The smoke and mirrors truly are there to distract, keep you from looking behind the curtains.
  • ““Those who cannot remember the past are condemned to repeat it.”

What’s next?

Read Full Post »

Arriving Saturday afternoon, I meandered over to the convention center that evening to pick up my badge, start examining the full conference guide, read breakout sessions’ abstracts and noted, like most conferences, that Saturday was partner sales day. Different vibe than when the users, attendees show up, but I digress. There was a full day of sessions on Sunday which you can see here, but this year I chose to pony up and spend the day in classes nosing around in virtual machines. First up, Pega 7 Overview 

This time last year in Orlando we were getting a preview of PegaPRPC 7 and the new Designer Studio. Pega 7 has the goals of:

  1. Works the Way the Business Thinks
  2. Simplifies and Accelerates
  3. mmm… I don’t remember. So much for the theory that doing, saying things in “3’s” will stick in your head. 😉

In any event the idea, or emphasis, is that Pega 7 is for building business rules for business users. As part of the introduction the deliberately provoking statement “Business process don’t change, they’re static” was made, but then the punch line elaboration was that it’s actually the business rules that are changing constantly. Back to the lab and Pega 7, Designer Studio is built on HTML 5. Beyond the initial layout when adding screen elements, moving things around is as simple as left-clicking and dragging them around to where you want them. Harnesses, flow actions and such are still there under the hood, but they’re not up in your face as in the past and Designer Studio puts a nice facade over that.

For those back in the PRPC 6.x world they do have a migration tool. With some advance planning it should be a fairly, if not fully, automated process. Patches are now fully inclusive as well. (e.g. – PegaPRPC 7.1.4 will include everything in 7.1.1-7.1.3).


On to Case Designer for Business Architects. I’m a technical guy at core, but when wearing my solutions or enterprise architect hat at a client site I like to see, know what the BAs are up to as, in recent years, multiple BPM platforms and vendors have been moving towards getting knowledge workers and business architects into the process designer to design the process flows.

A few definitions off deck slides in the class:

  1. “A case (n) is a business transaction to solve.” They have status and at least one process, have actors, tasks, data and history.
  2. “Process management (v): adapting to changing business conditions.”
  3. Case management (n): is “a holistic view of a business transaction” and a case type are the tasks needed to automate a business transaction.
  4. Steps are an action, a step in a stage.

So now we know what case management looks like from this vendor’s point of view. Diving down deeper we have the idea of “stages,” which would be key milestones, markers, phases that the business process must go through. Guardrails are still there  and the “rule of 7” still applies. That is, for the first-level grouping of stages in a case keep it to 7 +/- 2 and if you have more than that chances are you haven’t decomposed enough. When defining stages pay attention to the transfer of authority or a significant change in status.

Some quick best practices for defining case steps:

  1. use an iterative approach
  2. ignore details of each step
  3. set the expected order of tasks
  4. steps are universally understood
  5. limit decomposition levels
  6. easily communicated

Standard schtuff. In Pega Case Management every stage needs at least one step, the “Default step,” and the steps can be configured as single step assignments, multi-step or a case. In other words, case management is just a bunch of multi-step processes (a split join in BPMN parlance). SLAs and tolerance intervals are still there, of course, and can be set at the case level and the assignment level both.

A couple of other items of note are that PRPC can do federated case management, but the documentation says only between versions 6.3 <> 7.x. There was some talk of people in the UK doing it with versions 6.1 and 6.2, but that the exercise wasn’t trivial.

Also of note is that the instructor was accessing Designer Studio on a Mac with Google Chrome. Internet Exploder 11 isn’t yet compatible with PRPC 7.1.4, but will be. Conversely, Pega 6 doesn’t work with Chrome. The goal, direction is for there to be no asterisks on browsers and versions.

That’s what stuck in my head, I’ll doink around more on Pega Academy later and get lower to the ground on this stuff.

Props to Pega for putting the PRPC server in a small Ubuntu, Postgres VM and accessing the PRServlet URL from the host OS browser. Those puppies were small and fast (less than 7GB).

Read Full Post »