We’re in the Wednesday home stretch; this session didn’t have a specific theme but it seemed to mostly deal with unstructured processes and event-driven systems.
The session started with John Reynolds and Amy Dickson of IBM on blending structured flow and event condition action patterns within process models. John showed how they are modeling ad hoc activities using BPMN (rather than CMMN): basically, disconnected activities can have precondition events and expressions specified as to when and how they are triggered, be identified as optional or mandatory, and their behavior. It’s not completely standard BPMN, but uses a relatively small number of extensions to indicate how the activity is started and whether it is optional or required. The user sees activities with different visual indicators to show which are required or optional, and if an activity is still waiting for a precondition. This exposes the natural capabilities of the execution engine as an event handling engine; BPMN just provides a model for what happens next after an action occurs, as well as handling the flow model portions of the process. They’re looking at adding milestones and other constructs; this is an early pre-release version and I expect that we’ll see some of these ideas rolling into their products over the months to come. An interesting way to combine process flows and ad hoc activities in the same (pre-defined) process while hiding some of the complexity of events from the users; also interesting in that this indicates some of IBM’s direction for handling ad hoc cases in BPM.
Ashok Anand and R.V.S. Mani of Inswit presented their beta appiyo “business response platform”, which is an application development platform for small, simple BPM apps that can interconnect with social media such as Facebook, but an overly-short demo followed an overly-long presentation so difficult to grasp much of the capability.
We finished the day with Jason Bloomberg of EnterpriseWeb discussing agent-oriented architecture for cross-process governance: a “style of EA that drives business agility by leveraging policy-based, data-driven intelligent agents”. They call their intelligent agent SmartAlex; it’s like Siri for the enterprise, dynamically connecting people and content at the right time in a goal-driven manner rather than with pre-defined processes. Every step is just an event that calls SmartAlex; SmartAlex interprets models, evaluates and applies policies and rules, then delivers results or makes recommendations using a custom interface and payload depending on the context. Agents can not only coordinate local processes, but also track what’s happening in all related processes across an enterprise to provide overall governance and support integrated functions. EnterpriseWeb isn’t a BPM tool; it’s a tool for building tools, including workflows. Bill Malyk joined remotely to do the demo based on resolving a declarative conflict of interest; he showed creating an application related to cases in the system, and stating that potential conflict of interest cases are those that have relationships between people involved in the case. This immediately identified existing cases where there is a potential conflict of interest, and allowed navigation through the graph that links the cases and the criteria. He then demonstrated creating a process related to the application, which can then run flow-oriented processes based on potential conflicts of interest found using the declarative logic specified earlier. Some powerful capabilities for declarative, agent-based applications that take advantage of a variety of data sources and fact models, with greater flexibility and ease of use than complex event processing platforms.
My brain is full, so it must be time for dinner and another evening of drinks and conversation; I’ll be back tomorrow with another full morning and half afternoon of sessions.