Use of artificial intelligence (AI) tools in businesses has skyrocketed in 2024, with nearly three-quarters of surveyed companies using AI tools in at least one business function, according to research by McKinsey and Company.
From large language models (LLMs) to generative tools, AI has come to the forefront of technology discussions. It’s no surprise that everyone wants to leverage a tool that seems to work magic with so little user effort. However, it’s the “intelligence” part of the AI equation that needs to be carefully considered before entrusting valuable process management tasks to the technology.
One of the key challenges with using AI in process management is understanding the basis of its training. Any AI is only as good as the data it’s been built on. There’s no external ideas or information, no inspiration or creativity. It takes the material it has and arranges it according to instructions. If those instructions are to find patterns of behavior, that’s what it will do, but it won’t understand what those patterns really mean.
AI’s limitations in process mining
Unfortunately, a lot of the applications of AI in process management rely on exactly this approach. Process mining is a great example. An AI agent is unleashed on an organization’s internal systems to gather as many data points as possible, and then looks for patterns that it can raise to the attention of business analysts to suggest as process activity that could be captured, recorded and potentially improved through automation.
There are a couple of drawbacks to this methodology though. First, it still drops the real work of creating processes back in the lap of process experts. While the data is available to point to key elements of a process-in-waiting, it’s up to the human analysts to put that together in a workable process map that’s optimized for the organization’s operations. More importantly, if the entire system is predicated on finding certain activity types or key functions, it may miss real opportunities for process excellence. As it’s only able to work with what it understands as process actions, it can’t innovate or think outside the box of its training.
Making AI an artisan of process excellence
This isn’t to say that there is no place for AI in process excellence. Quite the opposite, it’s an exciting tool with a lot of potential if we direct it properly. I’d like to suggest that rather than sending AI bots into the wild to harvest possible processes, we instead make them artisans of process excellence and direct them to data that they can craft into something really effective.
This begins by changing the training approach. Rather than teaching AI to identify actions that could be part of a process, we need to train it on what makes a great process in the first place. That means teaching them principles of process writing – the importance of recognizing the everyday from the exception, how tasks should group into activities and how actions lead process steps through giving users a “how” as well as a “what.” By establishing sound process writing practices, we give AI a clear framework to build from.
Then we give it the building blocks. Your teams are already working steadily on tasks that should be captured, mapped and optimized. While traditional AI might identify some of these tasks, the people on the ground already know what they are.
Here’s where AI can be deployed effectively. Make the AI your staff’s process partner. With a simple instruction they can set the AI to examine their workflows and take in everything they do around a common task. Instead of combing through thousands of activities and hundreds of iterations, the tool can apply the best process practices to clearly defined examples and come up with concise documentation that opens the way to continuous improvement. Business analysts can go from trying to interpret bulk AI reports to the real work of optimizing process execution because the capturing and mapping has been done by their expertly trained AI assistants.
Freeing up process analysts
This is where AI can excel, and your teams can focus on doing what they do best – furthering your business goals. As they direct the AI to follow their day-to-day activities, the investment in process methodology training for the AI tools ensures that those tasks can be turned into solid process documentation without interrupting the work. Automation, optimization and continuous improvement become the focus for process experts who are freed from the demands of capturing the content at the ground level.
AI is going to continue to grow in prominence and effectiveness across every stream of business. Process management is no exception, but now is the time to consider the best ways to make use of the technology, rather than getting swept up in the hype and miss an opportunity to achieve excellence through it.