Building the Bench: Assembling an Interdisciplinary Team for Scientific Software Development

To build a better world with data, you need to be equipped with the right people.

Epiverse team at the Epiverse TRACE Summit. / Equipo de Epiverse en la Cumbre Epiverse TRACE.

The development of digital tools for social impact and specifically for scientific research and technically driven decision-making requires a new approach to professional development and team building. In Scientific Software Development (SSD), particular challenges must be faced, for “social impact products demand a more complex design process than commercial products do because of the complicated ecosystem of data providers, partnerships, and revenue streams required to bring them to life.” That complex design process depends upon a diverse team and one in which roles and skills are defined differently from those existing in the industry.

To face these demands, we should move away from conventional understandings of the social division of labor and consider SSD—and Open Source Software (OSS) at large—as an emergent field, with specific needs and changing dynamics. 

For instance, we know that collaborative networks are important for the sustainability of OSS projects. In other words, OSS projects rely on community participation for progress. This community-based growth of OSS highlights how people sustaining OSS collaborate, for interaction functions primarily occur through indirect communication (via platforms such as GitHub), even questioning traditional division of labor structures in this type of work.

How does the division of work operate in OSS and particularly in SSD initiatives when facing their challenges? 

This is not a simple question, but there are key themes worth exploring. 

During the Collaborative Software Development Ecosystem for Public Health – Epiverse-TRACE Summit, which was held in Bogotá from June 26 to July 7, 2023, members of the Epiverse-TRACE initiative met to discuss topics such as collaborative software development, team building, and community engagement. In these conversations, we posed three new questions:

  1. What roles are necessary for SSD enterprises and what does this word (role) mean in context? 
  2. What are the structures for collaborative work between different roles? and
  3. What kind of growth opportunities do these roles bring?

Context is Everything

When discussing the intricate design process involved in this type of software development, one could compile an exhaustive list. Indeed, in the summit discussions we identified a list of 20 different roles that can be tailored to the needs of the SSD and involves many more fields than software engineering for research and . However, it is worth highlighting that beyond Research Software Engineering and Data Science roles, those focused on the localized and contextualized engagement, design, testing, and direction of technological tools are equally critical to produce tools that could be globally connected and, at the same time, locally situated and significant.

For instance, domain expertise—such as epidemiology in the case of Epiverse—is crucial for orienting and defining SSD’s purposes. Domain experts represent the community of practice for which a software solution is produced, offering knowledge from a highly specialized user perspective. Social scientists, communicators, UX/IX designers, and training experts bring to SSD a user-centered and context-sensitive perspective to SSD, thus minimizing the risks of being lost in the translation between different styles of thinking and communities.

Simultaneously, traditional and new research coordination roles are necessary for sustaining the SSD initiatives, by mobilizing funding and promoting collaboration within and beyond projects. These roles play a crucial part in establishing new alliances with stakeholders and managing the human talent, financial, and bureaucratic resources required for the successful development and implementation of scientific software solutions.

In SSD, roles should not imply a strict division of labor but rather delineate collaborative spaces that facilitate the achievement of shared goals. 

Viewed from this perspective, we can understand how developers engage in translation activities throughout the development process by putting development outputs in tune with users’ definitions of priority, concerns, and needs. Similarly, social scientists and domain experts also do data science when reflecting on, and challenging different understandings of data and its uses, utility, and consequences.

Cultivating Collaboration Across Roles

As a relatively new field, we are encouraged to build new and open forms of work. If we understand roles in SSD as spaces for collaboration, what should this collaboration look like? SSD calls for a movement from rigid disciplinary efforts to forms of transdisciplinary cooperation.

On one hand, producing valuable SSD outputs requires different knowledge and communities. On the other hand, this knowledge is not limited to specialized professionals involved in a project. It is distributed among different types of users, including diverse external domain experts, other developers and collaborators who can interact with and improve SSD outputs, organizational stakeholders, and decision makers.

New Field, New Opportunities

Let us think of SSD as an opportunity. Placed at the interaction of the worlds of research and decision-making, SSD has the potential to avoid the traditional misunderstandings and uncontrolled mistakes that have fueled distrust and unease on both sides of this interface. SSD as transdisciplinary collaboration offers an opportunity to value the standpoint of different actors, to engage them in more open and humble conversations on decisive topics and translate this into durable technological outcomes.

Getting on the same page will help, but improved communication is not the only opportunity created by SSD.

As a transdisciplinary effort, SSD projects constitute an ideal environment for the career development of those involved. Building on Epiverse’s vision to place people at the center of software development, we can say that “the most important determinants of a successful OSS project are first and foremost about humans, and how they interact.” Cross-fertilization between different expertise, standpoints, and global connections open doors for personal and professional growth by providing the interchange of a wide range of skills.

As with skills, so with perspectives. Diversity in the broad sense is central to SSD enterprises. The engagement of different voices and experiences strengthen development outputs by improving their relevance and usability and fostering a sense of ownership within the user community. At the same time, it promotes an enabling environment for horizontal learning. This is not a minor thing if we consider how important it is for overcoming important barriers that are the business as usual in STEM fields, like the gender gap. By doing so, SSD shares the potential of different data science initiatives to contribute to a more equitable world.

The process of building diverse Scientific Software Development (SSD) teams highlights that SSD is more than simply writing code to develop innovative solutions. It emphasizes how the innovative nature and solution-oriented quality of technological advancements depend on a robust and diverse network of interactions among various roles, expertise, and collaboration mechanisms.

Stakeholders and funders should recognize this complexity when implementing and supporting SSD projects. Doing so will not only produce more responsible development outputs but also foster more diverse and inclusive development environments. Furthermore, it will enhance the professional growth of the individuals involved, who are tasked with transforming global and local infectious disease responses, while contributing to a more generous and sustainable world.

About the authors

Miller Díaz Valderrama

Qualitative Researcher for Epiverse-TRACE LAC

Universidad de los Andes (Uniandes)

Miller Díaz Valderrama is part of the sociotechnical characterization team for the Epiverse TRACE-LAC project at the Universidad de los Andes.

Read more

Laura Gómez Bermeo

Training Coordinator for Epiverse TRACE LAC

Pontificia Universidad Javeriana (Javeriana)

Laura Gómez Bermeo is the Training Coordinator for the Epiverse TRACE LAC project at Pontificia Universidad Javeriana. She is a Colombian mathematician with a master’s degree in education management and leadership from the UK.

Read more