Mónica Muñoz Torres - University of Colorado Anschutz Medical Campus
Yulia Levites Strekalova, University of Florida
Sara Singer, Stanford University
Mónica Muñoz Torres, University of Colorado Anschutz Medical Campus
Colleen Cuddy, Stanford University, School of Medicinie
Janani Ravi, University of Colorado Anschutz Medical Campus
Jake Y. Chen, University of Alabama at Birmingham
Sean Davis, University of Colorado Anschutz Medical Campus
Pamela Foster, University of Alabama at Birmingham
Jamie Toghranegar, University of South Florida
Modern scientific inquiry increasingly addresses phenomena of such vast scale that they strain the processing limits of individual researchers and traditional laboratory models. In the high-stakes domains of digital health and biomedicine, this paradox of scale reveals a systemic crisis where conventional, static research structures are ill-equipped to synthesize the resulting density of multimodal information. We propose dynamic teaming as a necessary strategic evolution of big team science, reframing collaborative interactions as the fundamental engine of emergent research cognition rather than a collection of individuals.
The dynamic teaming framework is architected around three interdependent pillars: people, data, and ethics. The people pillar moves beyond static organizational charts to leverage team scaffolds and multi-level decision-making, reducing cognitive load and managing distributed expertise. The data pillar treats information as a foundational architecture and external memory, employing coordinated convergence and FAIR principles to ensure datasets are AI/ML-ready across diverse domains. Finally, the ethics pillar serves as the collective's meta-cognition, embedding regulatory expertise from project inception to treat moral and legal considerations as proactive design constraints.
Drawing on implementation insights from the $130 million NIH Bridge2AI consortium, we illustrate how this methodology transforms operational frictions—such as leadership transitions, timing misalignments, and multi-institutional regulatory hurdles—into structured protocols for stability. Bridge2AI serves as a real-world test, integrating complex genomic maps, physiological waveforms, and voice biomarkers into a unified framework through these agile systems. By treating organizational complexity as a strategic asset, dynamic teaming enables large-scale collectives to function as unified cognitive systems capable of generating trustworthy, precision health breakthroughs. We conclude by calling for a systemic shift where funders and institutions recognize collaborative infrastructure as essential scientific infrastructure.