BPMN is Dead Now, AI Can Do it All

We've all seen how tools like ChatGPT can handle a variety of business tasks, automating nearly everything. And it’s true, Generative AI really can do a wide range of tasks that humans do currently. So, why bother designing and running your business processes with automation standards that were defined twenty years ago in a different technology era? Why not let your business users work directly with AI to do it all?

What was BPMN for?

The Business Process Model and Notation (BPMN), introduced in 2004, emerged from earlier standards like Business Process Execution Language (BPEL). Its purpose was to create a standardized way to model business processes that could be easily understood by all stakeholders—ranging from business users, managers, and analysts to software developers and the systems that executed those processes. Why was this so important? Because each group involved in defining the processes had different needs and expectations. Relying solely on a business user’s description of a task or activity wouldn't address the broader needs of the organization. For example, businesses must consider regulatory constraints—financial, privacy, security, and transparency requirements. Business users are not often well-versed in regulations like GDPR, so defining processes without input from legal, compliance, or security experts would lead to incomplete or flawed designs. BPMN ensured that all perspectives were considered and aligned.

Agentic AI takes the next step

The AI landscape is moving at an incredible pace, no longer just prompt-and-response. We're now seeing systems capable of reasoning over multiple steps and interacting with a network of AI “agents.” These advanced systems, known as Agentic AI or Large Action Models (LAMs), could eventually replace all the human roles involved in building business processes. Specialized AI agents, each trained for different business perspectives and expertise, could take over, eliminating the need for human intervention altogether. For example, the HR AI agent, interacting with the privacy AI agent, corporate-culture AI agent and systems AI agent would be able to define an appropriate employee-onboarding process.

But, how are these AI Agents going to be able to interact unambiguously? In reality, there’s little difference to when humans do it, because there still needs to be a way for the AI Agents to communicate between each other about processes. The quality and reusability of any result would be best if there was a sophisticated lingua franca for describing processes (more complex than a few simple steps). Well, that’s exactly what BPMN is. Crucially, one of its primary goals was that BPMN should be explainable to non-technical humans, so different people could validate what it would do. The way forward, then, is to make your AI agents generate explainable BPMN definitions of your business processes, possibly with external intelligent services contributing using the same common representation. Seven years ago, we demonstrated an example of AI trained on decisions in a process that gets dynamically converted to an explainable, executable, standard decision model notation, DMN:

Why use AI to run a process?

Once you have your AI-generated BPMN, you could then pass it to another AI Agent that’s been trained on how to execute BPMN, calling it each time a new instance of the process is needed. However, both training and running AI agents is computationally expensive and relatively slow, which makes it a real waste of resource (cost) when there already exists tried and tested, hyper-efficient and highly-scalable BPMN engines. It’s far more efficient, consistent and proven to run your BPMN on these specialized engines. Today’s BPMN systems should become the process runtimes for AIs. The same can be said for the other key business automation standards of Case Management Modeling and Notation (CMMN) and Decision Management Notation (DMN). Flowable has customers that run millions of BPMN and CMMN instances each day.

AI will love case management

As a final thought, now I’ve mentioned CMMN, there’s no better way to manage the execution of AI services and agents! Case management, as provided by CMMN, allows a way of expressing rich business automation that isn’t just about going through a sequence of steps to get to an end point. It allows a way of defining overall, end-to-end automation, with a 360° view of what may be appropriate at any point in the life-cycle of something, be it a person, a situation, a document, a project, or anything around which automation can be envisaged. It might decide when to request services and coordinate or discriminate their responses. CMMN has the concept of sentries (triggers) and stages (contexts) among other things. You can use contexts to define when certain AI agents are (or are not) appropriate, then have them activate given the right trigger. Depending on the results of those agents, additional AI agents or BPMN processes might then be triggered. All the while, human interaction or intervention can be included as part of the overall intelligent business automation. Five years ago, we called this approach intentional processes and highlighted the idea at BPMNext:

We will explore the leading role CMMN can have in managing real-world use of AI in more detail in a future post.