TeamFusion: Supporting Open-ended Teamwork with Multi-Agent Systems

2026-04-21

AI Summary

This work addresses the limitations of traditional answer aggregation methods in open-ended collaboration, which often overlook minority viewpoints and struggle to manage deep-seated disagreements. The authors propose a multi-agent framework that constructs preference-driven agents for each participant, explicitly surfacing consensus and dissent through a structured discussion protocol. By integrating preference-conditioned agent modeling, a formalized dialogue mechanism, and a consensus-oriented iterative synthesis process, the approach generates more balanced collective outputs. Evaluated on two collaborative tasks, the method significantly outperforms baseline approaches, demonstrating superior performance in both representativeness of individual perspectives and strength of consensus in the final outcomes.

Technology Category

• Multiagent Systems: Agreement, Argumentation & Negotiation
• Knowledge Representation and Reasoning: Preferences
• Humans and AI: Learning Human Values and Preferences

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Agentic search

Abstract

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.

Problem

• open-ended domains
• teamwork
• multi-agent systems
• answer aggregation
• minority perspectives

Innovation

• multi-agent system
• open-ended teamwork
• preference-conditioned agents
• structured discussion
• consensus synthesis