By Carlos Pessoa
For many years, governing processes essentially meant drawing the map.
This work involved bringing teams together, interviewing specialists and documenting each step through structured flows, using notations such as BPMN (Business Process Model and Notation) and EPC (Event Process Chain).
The result was consolidated into centralized repositories, organized and ready for consultation. However, there was a fundamental problem: these processes were rarely used.
In practice, while maps remained static — often “perfect” only on paper — day-to-day operations followed their own path, full of shortcuts, exceptions and improvised solutions that were never captured in flowcharts.
It was in this context that, around 2015, Process Mining emerged, bringing a new approach: transforming the digital traces left in systems into an accurate view of the real process, not just the idealized one.
From documented process to real process
With Process Mining, for the first time, it was no longer necessary only to ask how the process worked. It became possible to see it directly.
By bringing clarity to the real operation, even those who believed they knew the processes well were surprised to identify anomalies such as:
- Purchase orders created after delivery
- Approvals that skipped mandatory steps
- Suppliers paid without an associated purchase order
- Deviations between the planned flow and actual execution
In practice, Process Mining did not just bring answers. It brought much better questions.
When governance became a control panel
The natural evolution was to monitor not only how the process happened, but also whether it stayed within the expected limits of performance and compliance.
This led to the rise of near-real-time KPI dashboards, compliance alerts and maverick buying indicators — purchases made outside contracts that silently drain procurement budgets.
These advances allowed governance to stop being a dead archive and become a control panel.
Even so, there was still an important challenge: governance remained essentially reactive. The problem was only seen after it had already occurred. The failure and the record of non-compliance made the loss a done deal.
The arrival of AI in process governance
Over the last decade, governance has become a strategic priority, also driven by the adoption of Artificial Intelligence.
According to IDC, Brazil leads Latin America in AI investment volume, with expectations of reaching US$ 4.2 billion by 2026.
In the last two years, we have seen AI layers integrated into this observability foundation, generating a qualitative leap.
When AI is applied to Process Mining data, the scenario changes: the system starts to identify anomalies automatically and explain them in natural language. It stops working only as a report and begins to act as an advisor.
AI agents and the new phase of operational processes
Generative AI has opened a new chapter for process governance.
AI agents — systems capable of reasoning, planning and acting within operational flows — begin to interact directly with business processes.
In a scenario with thousands of purchase orders, for example, the technology can:
- Automatically identify price deviations outside contract terms
- Explain anomalies in natural language
- Recommend concrete actions
- Suggest renegotiations or volume redirection
- Support decisions before the financial impact becomes definitive
It is no longer just a report. It is an operational advisor.
The human factor does not leave the loop
Even with these automations, the human factor does not disappear. It is elevated.
Professionals begin to focus on higher-value activities, such as judgment, relationships and strategic decision-making, while repetitive execution is absorbed by agents.
In this new scenario, AI does not replace governance. It expands its ability to see, anticipate and act.
What will process governance become in this new world?
That is the question we must ask now.
What we are seeing points to a profound redefinition: governance in the age of AI is no longer focused only on maps, dashboards or non-compliance reports.
It starts to focus on real-time visibility into the actions of humans and agents, with the immediate ability to intervene, correct and improve the processes that matter most to the business.
This new scenario requires new instruments, new metrics and, above all, a new mindset.
From guardian of the flowchart to architect of dynamic systems
The process manager of the future is no longer just the guardian of the flowchart.
They become the architect of a dynamic system in which intelligence, automation and human judgment coexist in deliberate balance.
The map has always been a necessary simplification. What AI now offers us, for the first time, is the possibility of governing the territory directly.
This change goes beyond technology. It signals that it is time to apply the lessons of the past to explore the future.
About the author
Carlos Pessoa is Head of Delivery at Numen. He is a specialist in Process Governance and Digital Transformation, with a focus on the SAP ecosystem and Process Mining technologies applied to the CPG and Life Sciences industries.
About Numen
Operating in Brazil, Europe and North America, Numen is a consulting firm recognized for its expertise in SAP projects and as a strategic partner of major global players such as AWS, Salesforce and Celonis.
Recognized for its innovative approach and consistent focus on results, the company supports organizations in their digital transformation journeys, enhancing efficiency and generating sustainable value.
Learn more:
🌐 www.numenit.com
