Translating Data Science into Public Health Action: Lessons and Reflections on Academic–Industry Partnerships

Elena N. Naumova and Katie Stebbins, Tufts University, and Dantong Wang and Fabiola Dionisi, Nestlé Research 

Partnerships between academia and industry in nutrition and health data science are exciting but also inherently complex. At their best, they bring together complementary strengths: academic rigor and methodological innovation on one side and practical application, scale, and resources on the other. Yet these collaborations are shaped by a fundamental tension—academia seeks to generate generalizable knowledge, while industry is driven to produce timely, actionable, and consumer-relevant solutions. Rather than being a flaw, this difference is a defining feature, one that requires deliberate alignment if such collaborations are to achieve meaningful and lasting impact. 

In our experience, the tension often comes down to data. Access to rich proprietary datasets enables industry to contribute meaningfully to product innovation, yet open data sharing is logistically and legally challenging. While academia emphasizes openness, transparency, and reproducibility, the standards for data reporting and sharing are rapidly evolving. In fields of nutrition and health—where evidence informs both policy and public trust—questions of data access, ownership, and governance are especially sensitive. At the same time, advances in analytical methods, including artificial intelligence and synthetic data, introduce new hurdles related to interpretability and reproducibility. We’ve learned that without shared standards and clear agreements, even well-intentioned collaborations can struggle to balance the speed of innovation with time needed for verification and testing to ensure scientific credibility. 

Our experience has shown that these challenges are not only manageable, but they can become opportunities for innovation. Advances in methodological development are most impactful when they translate into accessible, practical tools for decision-making. A clear example is the collaboration between Nestlé Research and the Friedman School of Nutrition Science and Policy at Tufts University in developing the Global Nutrition and Health Atlas—a publicly available platform that integrates nutrition and health data for more than 190 countries, spanning more than 500 indicators across 26 topics and three decades. The GNHA exemplifies transdisciplinary, use-inspired research. It consolidates data from international agencies, NGOs, academic institutions, and peer-reviewed studies into six key domains—demographics, dietary intake, nutritional status, health status, health economics, and sustainable food systems—allowing users to navigate, analyze, and visualize information efficiently. We often think of GNHA as more than a data platform—it is a shared workspace where academic curiosity and industry experience meet. 

One comment we hear repeatedly from our industry colleagues is that platforms like GNHA are invaluable. Researchers rely on robust public health data to inform the development of science-based nutritional solutions across diverse populations and life stages. The rapid expansion of food and nutrition information calls for seeking new ways of data sharing and dissemination. Interactive platforms integrating data portals and visualization dashboards have been effectively used to describe, monitor, and track information related to food and nutrition. The process of collaboratively building the GNHA demonstrates a shared capacity to translate complex data into actionable insights—an essential component of innovation in both public health and industry settings. The project also pushed us to think differently about how data dashboards should be evaluated. Such experience helped us design a wide range of metrics and evaluation principles aiming to improve data standardization and harmonization, dashboard performance, and usability; broaden information and knowledge sharing among researchers, practitioners, and decision-makers; and accelerate data literacy and communication. 

Our collaboration soon grew beyond building a platform to joint scientific engagement and dissemination. Through a dozen presentations at national and international venues, including the Annual Meetings of the American Oil Chemists’ Society with invited talks on the potential of causal AI, the development of online surveys, and usability of synthetic data, we have addressed critical questions of reproducibility and methodological rigor. These activities reflect an ongoing joint effort that integrates academic depth with industry relevance. As we worked together on the narrative review of applications of AI tools in lipid nutrition fields, we discovered new opportunities for data analytics from nutrient interactions, lipidomics, and mechanisms at the molecular level to personalized nutrition, product development, and predictive modeling for public health at the global level. 

Along the way, we also learned several practical lessons for sustaining such collaborations. Synchronizing academic calendars with industry timelines and project deadlines requires careful planning, but it is entirely possible. Establishing clear legal agreements and expectations early on helps prevent misunderstandings and supports long-term trust. Integrating collaborative projects into graduate education creates opportunities for students to engage directly with real-world data and challenges while also learning about authorship, ethical responsibilities, and the iterative nature of research—from abstract submission to publication. These experiences foster a culture of open and transparent scientific practice. 

Looking back, the strongest evidence of success isn’t only found in publications or presentations. It comes from what collaborators tell us: better workflows, deeper understanding, stronger trust, and broader opportunities. These experiences also changed how we think about graduate education and how to better imitate real-life conditions by honing transferrable skills.  

Looking forward, we emphasize the need for building modular project-based courses, incorporating feedback and feedback-on-feedback, fine-tuning classroom teamwork, and engaging students in cutting-edge initiatives of the Tufts Food & Nutrition Innovation Institute. 

Our collaboration grew from a productive effort of teams at Tufts University and the Nestlé Research, supported by dozens of contributors, five generations of graduate students at the Friedman School of Nutrition Science and Policy of Tufts University, and three rounds of participants at the Health and Nutrition section at the annual meetings of the AOSC. 

Perhaps the most important lesson is that successful collaborations depend on a shared commitment to advancing science and serving the public good. The GNHA platform illustrates how jointly developed resources can serve a wide range of stakeholders—researchers, policymakers, practitioners, and industry partners alike. In addressing global challenges such as nutrition, health disparities, and food systems, solutions must extend beyond institutional boundaries. The lessons learned from such collaborations reinforce a broader insight: The goal is not to eliminate the differences between academia and industry, but to align them toward common purposes—advancing knowledge, informing policy, and improving health outcomes in a way that is both scientifically robust and socially responsible.