TeamFusion: Supporting Open-ended Teamwork with Multi-Agent Systems
For teams in open-ended domains, TeamFusion addresses the limitation of answer aggregation methods that suppress minority perspectives, enabling more consensual and representative team decisions.
TeamFusion, a multi-agent system, supports open-ended teamwork by instantiating proxy agents for each member, conducting structured discussions, and synthesizing consensus-oriented deliverables, outperforming direct aggregation baselines across metrics and tasks.
In open-ended domains, teams must reconcile diverse viewpoints to produce strong deliverables. Answer aggregation approaches commonly used in closed domains are ill-suited to this setting, as they tend to suppress minority perspectives rather than resolve underlying disagreements. We present TeamFusion, a multi-agent system designed to support teamwork in open-ended domains by: 1. Instantiating a proxy agent for each team member conditioned on their expressed preferences; 2. Conducting a structured discussion to surface agreements and disagreements; and 3. Synthesizing more consensus-oriented deliverables that feed into new iterations of discussion and refinement. We evaluate TeamFusion on two teamwork tasks where team members can assess how well their individual views are represented in team decisions and how consensually strong the final deliverables are, finding that it outperforms direct aggregation baselines across metrics, tasks, and team configurations.