Not so long ago, in the early 2000s – a time that now seems to belong to the pre-history of computing – the simultaneous arrival of industrial scanners and high-performance image encoding algorithms (JPG, PNG, with or without loss of quality) paved the way for a revolutionary idea: the dematerialization of documents and business processes.
At the time, the first workflow tools – still rudimentary state machines – offered a glimpse of a future in which business processes would be better structured and more fluid.
This promise came at just the right time. With the “baby-boom” effect, a demographic bombshell for the age pyramid, the mass retirement of boomers was jeopardizing back-offices, where manual task processing could no longer be carried out as before. Work methods had to be rethought.
However, despite the obvious potential for boosting productivity, the deployment of these solutions has been laborious.
Why was this? Because business units had little control over their own processes, which were often heterogeneous and ill-defined. And when these were finally mapped and integrated into workflows, all the inconsistencies and inefficiencies accumulated over the years became calcified.
The move to BPM: a first step forward
Faced with these limitations, software publishers came up with a solution: more flexible tools, accompanied by metrics to optimize digitized processes. The result was BPM (Business Process Management), a sort of workflow 2.0.
Business Process Management (BPM) is a systematic approach to improving an organization’s workflows by analyzing, designing, optimizing, and automating business processes. The goal is to enhance efficiency, reduce errors, and improve overall performance.
BPM often involves technologies like workflow automation, process modeling, and analytics to streamline operations. APIs enable BPM systems to integrate with various applications, automate workflows, and facilitate real-time data exchange.
These tools truly democratized the process approach, offering a more dynamic and adaptable vision. But a decade later, around 2015, a problem persisted: these digitized processes remained heavily dependent on human intervention at almost every stage.
Experts have even compared BPM to the three units of classical theater – time, place, action – a step could only change if one of these elements were to evolve.
So how do you automate these steps that are often characterized by manual input into customer software?
That’s where RPA (Robotic Process Automation) comes in. This technology has made it possible to code “little robots” capable of imitating human tasks, integrating directly via existing interfaces.
Robotic Process Automation, or RPA, works by deploying software bots that capture user actions, convert them into automated scripts, and execute tasks through the same user interfaces humans use—clicking buttons, entering data, and navigating applications—without requiring changes to existing IT infrastructure.
Bots can be scheduled, triggered by events, or manually started, while platforms monitor their performance through dashboards.
It’s not necessarily the most elegant solution, but it was effective. RPA sold like hotcakes, offering a pragmatic solution to a complex problem.
Agentic AI: the next disruption?
Today, a new wave of technology could turn everything upside down: agentic AI. By integrating with RPA and BPM tools, these artificial intelligences promise to go much further.
Agentic AI refers to artificial intelligence systems that can independently perform tasks on behalf of users with minimal human supervision. These AI agents can:
- Take initiative to accomplish goals
- Make decisions autonomously based on their understanding of the environment
- Execute complex sequences of actions across different applications or systems
- Learn from experience and improve performance over time
- Interpret instructions and figure out the necessary steps to complete them
What distinguishes agentic AI from other AI systems is its ability to operate with a degree of independence and purpose – rather than just responding to specific prompts, it can develop and execute plans to achieve broader objectives.
See also: The Rise of AI Agents: How Arazzo Is Defining the Future of API Workflows | The New Stack
In business process automation, agentic AI could automatically learn from human actions and analyze existing interfaces, then deploy intelligent robots in the blink of an eye to replace operators in repetitive, thankless business process tasks.
As for humans?
They could concentrate on high value-added missions, those requiring genuine expertise. But even these experts could soon be outdated, given the speed at which advanced language models (LLM/LAM) are evolving.
See also: Could LAMs Break the API Wave?
Tomorrow, the combination of agentic AI with RPA and BPM is likely to profoundly transform not only back-offices, but also customer interactions.
Imagine personal assistants powered by these technologies, integrated into corporate interfaces via well thought-out APIs. The companies offering the best openings and the most accessible systems will clearly be ahead of the game.
Related: Maximize Business Potential by Harnessing the Power of API as a Product
The realization of a promise?
Is the great vision of the 2000s – of comprehensive, seamless business process automation – about to become a reality? The next few months will tell, as things are moving very fast in this area.
In the meantime, a word of advice to companies: don’t delay the intelligent opening up of your information systems.
Agentic AI represents the next evolution in business process automation, but these autonomous agents require seamless access to high-quality data across organizational boundaries to function effectively.
This is where Axway’s strengths become invaluable; for example, our Amplify Platform, built on a federated API management foundation, creates consistent pathways for agentic AI to access and use data regardless of where it resides.
Businesses need to bring data and processes together from various systems – on-premises, in the cloud, or with external partners – to unlock AI’s potential.
Axway’s portfolio of integration solutions provides the necessary infrastructure to connect systems that were never meant to work together, enabling a unified flow of data.
Agentic AI is knocking at the door, and it won’t wait.
Let’s talk about it!
Learn how Axway enables AI-ready data across complex enterprises.
Follow us on social