As the tech sector pushes the use of artificial intelligence (AI) in the enterprise, there is an emerging trend which is seeing technology providers pivot towards the use of AI to orchestrate workflows.
This follows the agentic AI hype, but is focused on what analyst Forrester defines as “a clear shift from task-level automation to process orchestration for enterprise scale”. In Forrester’s The adaptive process orchestration software landscape, Q2 2026 report, published in April, the analyst noted that the market is maturing around agentic and AI-first approaches with what it sees as “an emphasis on blending adaptive AI behaviour with deterministic workflows rather than entirely replacing structures”.
Its research found that software providers in this market tend to highlight consolidation of automation tools into orchestration backbones that combine process intelligence, modelling, execution, monitoring and data foundations. According to Forrester, there is also a strong emphasis on governance, auditability and hybrid execution models that support event-driven automation and human-in-the-loop.
Camunda, one of the companies covered in the Forrester report, held its annual conference this week (19-21 May) in Amsterdam. With 1,100 attendees, there appears to be a lot of interest in the use of agentic AI, to enable businesses to run more efficiently by reengineering their business processes. Opening the CamundaCon event, the company’s CEO, Jakob Freund, discussed why a safe customer data agent requires a high level of human approval, and needs to be deterministic. He described these traits as “the power of agentic orchestration”, which enables organisations to use agentic AI safely.
Freund referenced the Forrester report, adding that every process in an organisation is legacy, as it was designed at a time when AI did not exist. He suggested that AI completely redefines how organisations operate.
The keynote presentation was used to show delegates the approach Camunda is taking internally to reinvent its own business processes using a product launched at the CamundaCon called ProcessOS.
During his presentation, Camunda chief technoogy officer Daniel Meyer described ProcessOS, as “an agentic operating system” that he said reengineers business processes and then continuously optimises them for the world of AI. At Camunda, he said ProcessOS was used to redevelop the quote to cash process. “It is one of our most critical business processes, and there were quite a few inefficiencies, manual handoffs, spreadsheets,” he said, adding that once it was reengineered using ProcessOS, Camunda was not only able to improve the cycle time and efficiency, it also reduced the error rate.
Camunda’s chief financial officer, Clemens Morgenroth, said the reengineered process is freeing 6,000 person hours, based on the fact that the quote to cash process used to take five hours per deal.
Orchestration of agentic workflows at Barclays
Barclays is among the companies looking at reengineering business processes around agentic AI using the Camunda platform. Lily Wang, CIO for wholesale client onboarding and group financial crime at Barclays, said: “ProcessOS tackles the real reason AI adoption stalls in large enterprises. We can’t build tomorrow’s process using only what we know today. Transformation stalls.”
During a presentation looking at how Barclays uses agentic AI to augment the onboarding of new customers, Gautam Verma, the bank’s head of financial crime core platforms and client due diligence technology, discussed the use of deterministic and agentic orchestration.
“The complexity of the customer due diligence process can get very involved and can take months to do,” he said. “There’s a lot of sequential handles between peoples and teams that do different functions on this journey and the evidence gathering can be manual in a lot of cases.”
Verma said the bank started developing an agent that would help do data collection from multiple sources. It also introduced a data intelligence agent, which takes the customer data collected and assesses the policies and procedures that need to be taken for effective due diligence. There is then a third agent to handle the policy procedures themselves.
The idea is not only to streamline a business process but, according to Verma, to “fundamentally reimagine how we move around the customer onboarding process”.
He said the orchestration layer across this workflow is deterministic. “We’ve got multiple systems at play during this whole journey – multiple handoffs for the internal systems and internal teams – and we needed to orchestrate it completely end-to-end to achieve the goals that were set for us,” added Verma.
