First Low-Code breathed new life into BPM. Now AI (artificial intelligence) is pushing BPM back to the forefront.
I recently bought a ceiling fan from a well known online vendor. Before it arrived, I ordered an extension rod, having concluded that I would need it in order to mount the fan on a sloped ceiling. When the fan arrived, I quickly figured out that I had ordered the wrong unit, primarily because the online catalogue didn’t make it clear which version I needed.
With this particular vendor, the automated return process required a written explanation of why I was sending the item back. I typed into the field provided, “The fan did not work for my intended purpose.”
When I clicked the submit button, I was informed that I would not be charged for return shipping and I was then given a UPS shipping label to print out, along with instructions for how to repackage, etc.
A day later, when the extension rod arrived, I went through the same automated process. When asked why I was returning the rod, I wrote “Having returned the ceiling fan, I no longer need the extension rod.” I was then informed that I would be charged $6 for return shipping and provided a shipping label etc.
Contextual Analysis (AI) Triggers Process
Clearly, this particular online vendor is using Contextual Analysis (CA) software—a form of Artificial Intelligence—to extrapolate a trigger for an automated business process. CA is able to analyze unstructured data—in this case, writing—and understand its meaning, at least, to some degree. My first return explanation (the fan) included the words “did not work,” which caused the CA engine to conclude the fan was somehow defective and, therefore, that I should not be charged for return shipping. My second explanation had no such language, causing the CA engine to conclude that I should be charged shipping.
While this vendor’s return process is a simple example, it, nonetheless, illustrates the convergence of artificial intelligence (AI) and business process (BPM), one of the hottest new areas for high-tech investments.
BPM and AI Deliver Escalating Efficiency
In a recent report (“Artificial Intelligence Revitalizes BPM . . .”–subscription required), Forrester Research analyst, Rob Koplowitz points to this type of Artificial Intelligence (Natural Language Processing/Contextual Analysis) as one of the primary intersection points between AI and BPM. Another such intersection involves machine learning, the ability of AI engines to assess the relative efficiencies of automated processes and to make recommendations for modifications to business logic and even processes, themselves, in an effort to continually escalate process efficiency:
Not surprisingly, many BPM vendors focus on AI as a means to reduce the complexity of tuning and optimizing the very processes the systems manage. While the idea of using analytics to monitor process execution is nothing new, AI can now take that a step further through the use of machine learning to provide guidance on optimization.
Applying AI to BPM (process efficiency) is definitely a step in the right direction, but real-time implementation of process modifications is the next frontier.
AgilePoint NX, the Responsive Application Platform
AgilePoint, a long-time player in the BPM industry and quick leader in the Low-Code space, is often billed as the “Responsive Application Platform,” a distinction resulting from a unique architecture that enables applications (automated processes) to adapt at runtime to changing business and technical conditions.
AgilePoint’s model-driven architecture, an abstraction layer that translates model metadata in real time into functioning code, and a stateless process engine combine to yield this self adaptive capability. In regard to AI-driven process-efficiency analysis, rather than make recommendations to human developers, who must then refactor applications according to AI instructions, those same instructions can be fed directly into AgilePoint applications, which can adapt at run time to the continuous feed of process analysis.