Project Management & Agile··1 sources verified
Scrum Practitioner Argues AI Agent Teams Need Coordination, Not Just Better Models
Summarised by Relevant News AI · Read time: 3 min
The author argues that the 'Agent = Model + Harness' formula works for a single AI agent but fails for teams, where defects come from handoffs, shared records and duplicated work. He proposes 'Team = Agents + Coordination' and maps coordination onto Scrum, with the Product Owner and Developers handling orchestration and Scrum artifacts serving as shared state. He draws on his own team of AI agents, which runs on Scrum, to show where its defects originate.
Why it matters: Teams building multi-agent workflows can apply established Scrum coordination practices, such as backlog ordering, shared artifacts and timeboxed events, to fix failures that a better model will not solve.
All sources