cross-functional collaboration

This paper addresses the multi-skill expert team formation problem: given a set of tasks requiring collaborative, multi-skill execution, the objective is to maximize skill coverage across tasks while minimizing the maximum expert workload, with an extension to incorporate collaboration efficiency. We formulate— for the first time—the joint optimization of skill coverage quality and workload fairness as a multi-objective integer programming problem. We propose a theoretically grounded greedy heuristic algorithm with provable approximation guarantees. Our approach integrates integer linear programming modeling, multi-objective optimization design, computational complexity analysis, and empirical evaluation on real-world datasets. Experimental results demonstrate that the algorithm improves skill coverage by 12.7%, reduces workload variance by 38.4%, and achieves two orders-of-magnitude speedup over baseline methods—effectively balancing task completion completeness and team workload equity.

This work addresses the imbalance between structure and flexibility in existing multi-agent collaboration frameworks powered by large language models, which often leads to inefficiency and resource waste. To overcome this limitation, the authors propose LATTE—a novel framework inspired by distributed systems—that introduces a dynamically evolving, shared task graph to explicitly encode subtask dependencies, agent assignments, and progress states. By integrating LLM-based multi-agent systems with distributed coordination protocols, LATTE enables adaptive collaboration structures, dynamic task allocation, and emergent task discovery. Empirical results demonstrate that LATTE significantly reduces token consumption, execution time, and communication overhead across diverse collaborative tasks, while minimizing file conflicts and redundant outputs. Moreover, it achieves accuracy on par with or superior to strong baselines such as MetaGPT.

Multi-unit organizations face a persistent trade-off between unit performance improvement and cross-unit brand/behavioral alignment. Method: Using agent-based modeling, we simulate multiple interdependent organizational units with task similarity, examining how communication network topology, knowledge-sharing norms, and conformity behavior jointly affect both performance and coordination. Contribution/Results: We identify that moderate decentralization—specifically small-world or modular network structures—enables simultaneous enhancement of unit-level performance and organizational coordination, thereby transcending the conventional performance–alignment trade-off. However, under high inter-unit dependency, centralized control remains more effective for achieving coordination. These findings delineate the boundary conditions under which decentralization fosters coordination, revealing that structural decentralization alone is insufficient without appropriate task interdependence and normative scaffolding. The study advances a novel design principle for multi-unit organizations: balancing autonomy and coherence through context-sensitive network architectures and institutional mechanisms.

This study addresses the systemic support of macrocognitive functions—namely, event detection, sensemaking, adaptability, perspective shifting, and coordination—in human–AI teaming, moving beyond traditional usability-centered design paradigms. Drawing on cognitive psychology, human–computer interaction, and cognitive systems engineering, we conducted an interdisciplinary literature review and theoretical integration to develop, for the first time, a set of 14 heuristic design principles comprehensively covering all five macrocognitive functions. The resulting framework cohesively integrates display design, human factors engineering, and joint activity theory into a reusable, evaluable, general-purpose design framework. Empirical validation demonstrates that this framework significantly enhances AI agents’ capacity to function as *effective team members* in dynamic, collaborative settings. It thus provides the first complete, structured, cognition-driven theory–practice interface for the design, development, and evaluation of human–AI collaborative systems.

In software engineering, manual extraction of goals from stakeholder interviews and subsequent goal modeling suffer from low efficiency and poor reproducibility. To address this, we propose the first end-to-end automated goal modeling method integrating textual entailment reasoning with a large language model (GPT-4o). Our approach directly generates structured goal models from unstructured interview transcripts, supporting high-level goal-to-software-operation mapping, requirement refinement, and conflict/obstacle analysis, while enabling goal provenance tracing and refinement relation inference. Evaluated on 15 cross-domain interview datasets, it achieves a goal matching rate of 62.0% (comparable to human performance), a provenance tracing accuracy of 98.7%, and a refinement relation generation accuracy of 72.2%. The core innovation lies in the first application of textual entailment to goal modeling, significantly enhancing both the accuracy and interpretability of automated goal modeling.

Traditional software project management struggles to address the emerging paradigm of AI-driven Software Engineering 3.0, particularly in human-AI collaboration and the establishment of ethical and accountability mechanisms. This work proposes an “agentive project management” framework, introducing the novel concept of an “agentive project manager” and designing four autonomous operational modes that dynamically balance automation and human oversight. It redefines the human role as a strategic leader and AI agent coach. Built upon a multi-agent system architecture, the framework integrates tunable autonomy, autonomous decision-making, and human-agent collaboration to enable dynamic task adaptation and traceable accountability. This study lays a theoretical foundation for project management in the SE 3.0 era and outlines a systematic research roadmap to advance paradigm evolution and community development.

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.

This study investigates the impact of hybrid work arrangements on productivity and collaboration within agile teams, with particular attention to key challenges such as diminished informal interactions, unequal participation, and heightened reliance on digital tools. Drawing on semi-structured interviews and qualitative content analysis of three Norwegian agile teams, the research elucidates how communication patterns, collaborative mechanisms, and agile ceremonies function in hybrid settings. It uniquely highlights the mediating roles of trust, communication, and tool support, revealing that while agile ceremonies serve as critical alignment anchors mitigating collaboration breakdowns, hybrid work significantly curtails informal exchanges and exacerbates participation disparities. The findings underscore the necessity of tailoring team structures and enhancing digital tooling to foster inclusivity and sustain long-term performance.

This study addresses the persistent challenges impeding the effective integration of Agile and DevOps practices, which are often constrained by cultural, organizational, procedural, and technological barriers that undermine software delivery performance. Through semi-structured interviews with six senior practitioners from Brazil and Germany, the research employs qualitative thematic analysis to systematically identify—within a cross-national context—four core integration challenges and proposes a corresponding solution framework. The findings underscore the pivotal roles of cultural alignment, team autonomy, process coordination, and infrastructure automation, highlighting that organizational and cultural factors are critical enablers of successful technical integration. By elucidating these interdependencies, the study offers actionable, cross-cultural guidance for software organizations seeking to enhance their Agile–DevOps convergence and overall delivery effectiveness.

This work proposes Visual Milestone Planning (VMP), a novel approach that addresses the lack of intuitive, collaborative milestone planning mechanisms in hybrid development environments where agile teams struggle to integrate with traditional planning paradigms. VMP innovatively combines a milestone planning matrix with a physically inspired visual scheduling mechanism: product backlog items are mapped to milestones and arranged as Tetris-like work packages on a resource–time canvas, enabling dynamic determination of milestone deadlines. By bridging agile practices with conventional project planning, the method significantly enhances team collaboration, planning transparency, and shared understanding of delivery cadence.