Agentic workflows are great at deciding what to do next—until reality interrupts with payments, reservations, state changes, exceptions, and approvals. At that point, you need more than clever routing: you need transactional guarantees, compensation, fault recovery, and auditable, human-in-the-loop control across distributed services.
Are we reinventing yesterday’s orchestration layer by wrapping everything in tools… again?
This talk argues that workflow management is not “legacy”—it’s the missing layer for operationalizing multi-agent systems. We’ll show how a visual, no-code approach can make agentic processes observable and governable, while preserving data consistency in long-running, stateful executions.
You’ll leave with a clear blueprint for building agentic applications that are fast to develop, safe to run, and trustworthy to operate.

Agentic development for transactional, auditable enterprise workflows