Inhalt
The pressure to adopt AI is real, but handing steps to an AI agent without a clear design for human oversight is how workflows break down quietly. In this talk, I'll show how to architect AI workflows that keep humans in the right place at the right time: in control without becoming a bottleneck, and efficient without giving up accountability.
Using a single end-to-end scenario (draft, validate, publish) I'll walk through two workflow architectures: one where every step is defined and auditable, one where AI navigates the path itself. You'll see how the choice between them changes where human judgment is required, what guardrails are needed, and how you measure success. Whether you're a writer looking to work smarter, a team lead designing a new process, or a decision maker evaluating AI adoption, you'll leave with a framework you can sketch on a whiteboard and a checklist you can act on next week.
Das lernen Sie
After this session, attendees can: place humans correctly in AI workflows, choose the right architecture, and explain the trade-offs to stakeholders.