Agentic AI Spurs Process Rework at Camunda, Barclays

Camunda introduced ProcessOS at its CamundaCon conference in Amsterdam (May 19-21). The company reported that reengineering its quote-to-cash workflow freed about 6,000 person-hours, based on a prior average of five hours per deal.

Camunda CEO Jakob Freund argued that customer-facing AI agents must have high levels of human approval and deterministic behavior, calling those traits ‘the power of agentic orchestration.’ CTO Daniel Meyer described ProcessOS as ‘an agentic operating system’ that reengineers business processes and continuously optimizes them for AI-driven workflows.

Camunda CFO Clemens Morgenroth tied the 6,000-hour estimate to the earlier five-hour processing time per deal, and Meyer said the overhaul removed manual handoffs and spreadsheet work that had slowed processing.

Barclays is piloting agentic orchestration on Camunda’s platform to speed customer onboarding and financial crime checks. Lily Wang, CIO for wholesale client onboarding and group financial crime at Barclays, warned that organizations cannot design future processes using only current assumptions and said ProcessOS addresses reasons AI adoption stalls in large firms.

Gautam Verma, head of financial crime core platforms and client due diligence technology at Barclays, outlined a three-agent approach: an agent to collect data from multiple sources, a data-intelligence agent to assess policies and procedures, and a third agent to execute required policy steps. Verma said the bank’s orchestration layer is deterministic to coordinate multiple systems and human handoffs and to meet end-to-end goals.

An April Forrester Q2 2026 report cited at the conference described a shift from task-level automation to process orchestration at enterprise scale. The report said vendors are consolidating automation tools into orchestration backbones that combine process intelligence, modeling, execution, monitoring and data foundations, and that enterprises are emphasizing governance, auditability and hybrid execution models that support event-driven automation and human-in-the-loop controls.

CamundaCon drew about 1,100 attendees. Presentations and vendor messaging at the event highlighted internal use cases, governance requirements and the role of deterministic orchestration when deploying agentic AI in regulated industries such as banking.