It has been proven that end-to-end or large-scoped processes can behave intelligently by calling out to big data analytics, predictive analytics, cognitive knowledge services, mining, optimization, and signal or pattern recognition services. In many cases, the process software can embed these capabilities in a process/case to augment a case or process, identify opportunities for change, and make adjustments in flight. These are centralized and encompassing processes, but the idea of smaller sub-processes, or process snippets, that can be leveraged in different contexts in a static or dynamic mode is emerging. They may be either in a centrally controlled case or process, or acting autonomously. How do snippets get smart?
Called at the Right Time
Snippets are effective when they are linked to a case or process, and are often called sub-processes, which are created to take some form of action, like analysis or work tasks. Snippets are linked sets of associated services that work in a packet for the good of the central process or case. Often, they are reused and leveraged in multiple processes and elevate the granular service approach to associated and collaborative services.
Triggered by Signals or Patterns
Snippets can be context sensitive when they are linked to signals, event data, and patterns. Patterns are a set of signals, or event data, that are saved and watched over a period of time. By linking these snippets, which may convey a need for further analysis or action, they can lay dormant until the need of their work and outcomes is made apparent. This is a fast-growing approach for known signals, events, and patterns.
Focusing in on an Emerging Need
Snippets are enabled to look for opportunities instead of being locked to a process, case, or a signal/pattern, but the work they were designed to complete can occur without central control. These snippets sense for the right conditions to wake up and perform their duties. In advanced uses, they can swarm to a problem to bid on doing the work, or complete the work automatically. In this instance, the snippets are generally goal driven and constrained from bad behavior
Just like dynamic pixels make up a picture or moving videos, the process of the future will consist of multiple snippets instead of a single (or a few) monolithic process(es). These snippets will sometimes work in a process or case framework, or other times outside this framework acting independently and guided by constraints.