Patterns and Themes in RCD Professional Career Paths

Patterns and Themes in RCD Professional Career Paths Samuel Porter University of Maryland [email protected] Patrick Schmitz Semper Cogito [email protected] Deb McCaffrey Arizona State University [email protected] Matthew Humphreys Arizona State University [email protected] Kerk Kee Texas Tech University [email protected] Abstract Research Computing and Data (RCD) professionals play a crucial role in supporting and advancing research that involves data and/or computing, however, there is a critical shortage of these RCD professionals and organizations face challenges in recruiting and retaining RCD workforce roles. Many potential RCD professionals are unaware of these roles and how their skills and experience map to the RCD workforce needs. The CaRCC RCD Career Arcs working group has been working to gain a deeper understanding of RCD career paths. Following up on our survey of key factors that influence decisions in the course of their careers, we conducted 14 interviews to more deeply explore career choices, influences, and factors that help shape RCD career paths. This paper summarizes the findings and presents observations and recommendations for both hiring managers as well as current RCD professionals. Keywords— Research Computing and Data, RCD, Cyberinfrastructure, CI, Career Paths, Professionalization 1 Introduction “I didn’t know these jobs existed.” This is a quote from a recent candidate who applied for a Research Computing and Data (RCD) role at one of our institutions, and it echoes the sentiment often shared by colleagues: they never expected to become an RCD professional – they didn’t know these roles existed in the first place! Finding people to fill RCD roles at all levels continues to be a challenge for managers. While terminology and HR frameworks for these roles are being adopted[4], there is still much progress to be made in terms of understanding how best to recruit, retain, and support RCD professionals[1]. RCD infrastructure and services are essential to the modern research enterprise and continue to evolve and expand at a rapid pace. Researchers are struggling to keep pace with the explosion of data and the rapid evolution of computing resources and often lack the skills to make full use of emerging tools and techniques. RCD professionals provide the support researchers need to take advantage of the latest tools and services [11]. The CaRCC Career Arcs Working Group was formed to investigate RCD career paths and factors influencing career decisions. Several workshops have identified the importance of understanding RCD career paths. The 2017 NSF-sponsored Professionalization in Cyberinfrastructure workshop [2] emphasized the scarcity of RCD professionals. The findings of a 2020 workshop, Building the Research Innovation Workforce [10] emphasized “...there are few well defined and stable career paths” and “...it is not clear what career paths are available or how a person can pursue advancement.” A broad survey of RCD stakeholders[6] found that the issue “Documenting and supporting aspirational career paths for RCD professionals within higher education” was rated as both very important and quite difficult. Research by the CaRCC RCD Professionalization Working Group[8] demonstrated the wide range of starting points and ultimate roles for RCD professionals. They also note the need for clear career paths and a better understanding of how to retain and develop RCD professionals. Other research efforts have identified the importance of mentoring and professional development[3, 9]. Filinger et al.[7] discuss the importance of learning pathways and the need to help early-career RCD professionals understand the roles and the needed skills to enter RCD careers and be successful. Hacker et al.[10] also noted the lack of role models and mentors within the community. The Career Arcs Working Group’s earlier work[5] explored the key factors that RCD Professionals considered in making career decisions. Our current work follows up on that with 14 interviews of RCD professionals at various career stages. Analysis of the resulting data further explored commonalities in how RCD professionals found these roles, how they grew into them, and what factors and influences impacted their careers over time. This paper summarizes our approach to this work, what we learned, and some recommendations for both potential RCD professionals as well as managers hiring for these roles. ©2025 CaRCC licensed for use under Creative Commons Attribution-NoDerivatives 4.0 International (CC BY-ND 4.0) license. 2 Methodology This qualitative study used semi-structured interviews with 14 diverse RCD professionals to explore their career trajectories. Conducted via Zoom between May 2023 and February 2024, the interviews examined educational and professional backgrounds, pivotal career moments, motivations, challenges, and future aspirations. Grounded theory methodology[12] guided the analysis through open, axial, and selective coding, grouping themes into three career stages (pre/early, mid, and advanced) and four central topics (decisions made, motivations, factors and influences, and risks taken). Participants’ demographics varied in gender, race/ethnicity, age, and professional backgrounds, with most employed in academia or national labs. The study reached theoretical saturation and produced a career trajectory model consistent with prior research, offering rich insights into how RCD careers evolve through key decisions, motivations, and risk-taking behaviors. 3 What We Learned 3.1 Entering the RCD Field The stories of those RCD professionals who began on an academic career path demonstrated that many entered RCD as a function of opportunity. Most did not intentionally seek out their first RCD role, but were introduced to it by a mentor, or found it while looking outside of academia due to a lack of attractive academic opportunities (e.g., they did not look forward to the pursuit of grant funding and tenure competition but wanted to remain engaged in research). One said: “I stumbled into someone who showed me a position in research computing that I felt I could do while looking for more full-time employment. I found that I liked it a lot!” Some RCD professionals transitioned from industry roles, tech startups, manufacturing, enterprise IT (including campus IT), or corporate research labs. These career shifts were motivated by changing values (not supporting the mission of their company anymore), a desire to contribute positively to society, or a desire for a “more relaxed” work environment. While a PhD is not required for most RCD roles, many participants reported that having a PhD and/or experience with research, including a track record of publications, helped in interactions with researchers, who appreciated their understanding of the research methods. While experiences vary by institution and role, many people in RCD roles are equally successful without a PhD. 3.2 Importance of Mentoring, Community, and Network In earlyand mid-career stages, mentors (both formal and informal) and communities were described as significant factors in RCD career paths. By the same token, advanced-career participants expressed a desire to mentor and coach others. Often, mentors or supervisors were responsible for introducing participants to the RCD field. Participation in communities of practitioners – whether at their institution, in their region, or nationally – led to an increase in confidence about skills, creating a sense of belonging, and impacting career decisions, including whether or not to continue in the RCD field. These findings underscore the importance of efforts in the RCD community to expand mentorship and coaching services, including the new CaRCC Mentoring Working Group1. 3.3 Motivations, Factors, and Risk Tolerance vary by Career Stage Comparing factors and motivations that influenced decisions across career stages, we found a major shift between early-career and mid-career. Early-career decisions were often based on factors like financial necessity, career path dissatisfaction, or a lack of opportunity for advancement. These factors align with a general willingness to tolerate more risk, e.g., changing career paths, going back to school, or taking on roles with uncertain prospects. However for mid-career professionals, factors shift toward being interested in the work, an ability to have influence, and salary. In both midand advanced-career stages, participants described an increased focus on personal priorities, meaning job changes were motivated by personal goals and not necessity. Motivations for change included the right work environment (sometimes getting away from politics or a shift in mission), a desirable work location, and desirable circumstances for the role, such as building a new team versus managing a well-functioning group. Advanced-career factors included impact at a larger scale, taking on new challenges, and satisfaction in coaching and mentoring. As they advance, people choose the role they want, not the job they need, although these decisions are often balanced by increased risk-aversion due to family and financial constraints. Of note, when asked to imagine their future selves, most participants indicated they expected to remain in RCD roles. 1https://carcc.org/mentoring-working-group/ 2 In a common mid-career choice, we saw participants self-select into or away from management roles reflecting their evolving self-awareness of an inclination to manage others or to be an individual contributor. Some midand advanced-career participants described reaching the end of a promotion path at their institution, prompting a move to find new opportunities. This is particularly evident in management and leadership roles for which there are fewer positions, but also occurs in non-supervisory roles when institutions do not have well-defined paths for advancement. At the same time, other participants described a strong sense of loyalty and commitment to their institution and said they wouldn’t leave their position without completing in-progress projects. 3.4 Common Traits The RCD professionals we interviewed are passionate about some aspects of their role but we found that their passions shifted over time, often in line with shifting roles. For example, one participant expressed a passion for programming and computers that kept them in an RCD role early in their career (and literally in a lab). Now well into mid-career, a clear passion is coaching and developing people. Other examples we heard participants talk about were how much they like a given domain (e.g., bioinformatics or chemistry), solving problems, leading teams, or providing a service to researchers. One participant explained: “Having a job where I get to interact with research and support it from a service mindset is really rewarding.” We find that these traits reflect a reality that RCD professionals often face when providing support for research with a significant amount of scopeand resource-ambiguity. Participants described themselves as self-directed, passionate, curious learners who are willing to ask for help, share, and be open to new ideas. Though these traits don’t guarantee success or reflect all people who are in RCD roles, they represent a common thread running through many of the interviews. 3.5 Recommendations for Managers 1. Consider prioritizing efforts to connect new RCD professionals to mentors and peer RCD professionals, and encourage (and allow time for) involvement in RCD communities. 2. Recognize that early-career researchers are often taking significant risks to enter the RCD field and provide support for this change. 3. Managers should engage their institutions along with the RCD community to better define HR roles and career paths to make RCD paths seem less risky. 4. When recruiting and hiring, communicate what RCD professionals find rewarding about their roles and emphasize traits such as curiosity as much as or more than technical skills. 5. Early-career professionals are good candidates for RCD roles, and hiring managers should consider a range of fields/backgrounds. Managers must describe career paths, and may want to connect prospects with current RCD professionals who can advocate for RCD roles. 6. When recruiting and hiring midand advanced-career RCD professionals, keep in mind that they may be more risk-averse and less willing to relocate, but will often make lateral moves to further personal goals. By the same token, employers can retain these professionals by focusing on what they find fulfilling. 3.6 Insights for New and Current RCD Professionals 1. Participants from many fields and backgrounds found success and fulfillment in an RCD career, not just professionals with a computing background or an academic background. Notably, a PhD is not required, but professionals with a PhD or research experience said they felt better equipped to understand the researchers they support and are satisfied by the work. 2. Participants frequently described the importance of mentorship and their involvement in communities of practice for their career development and enjoyment. 3. Many participants advanced within their institutions by taking on supervisory roles, but there are also opportunities to advance for non-supervisors, such as technical or domain expertise (leadership as distinct from management). Some RCD professionals found success through self-advocacy, or by moving across institutions into roles that fit their career aspirations, particularly when the roles are not supervisory in nature. 4. This field is new and many institutions are investing in these roles for the first time. You can influence your own growth and may be able to define your own role. 5. Very few participants wanted to leave the RCD field; people like these jobs. Those that left generally did so due to changes in the work environment and not because of the work itself. 3 3.7 Visualizing RCD Career Paths In addition to the written analysis, this phase of work included the goal of producing a visualization to help RCD professionals understand the career paths open to them and to help managers understand sources for hiring and how to support existing RCD professionals with their professional development. We created a visualization inspired by transit maps, and submitted as a poster to PEARC25 [in submission]; the full image of the poster is available at: doi.org/10.5281/zenodo.15048236. Figure 1: Excerpt of the RCD Career Paths Visualization. 4 Conclusions and Future Work We set out to explore factors that affect RCD career paths, and to identify emergent themes around the interactions between those factors. Despite our small sample size, our work elucidated much in service of these goals. Our work indicates that RCD career paths are largely considered as alternative career paths: found by academic careerseekers who deviate from a traditional academic career path or industry career-seekers who feel dissatisfied with their industrial environment. We also found that an individual’s career stage strongly affected their motivations for taking on new roles, and their approach to risk. A lack of awareness and aspiration to RCD roles means managers will need to be thoughtful when describing, hiring, and recruiting for RCD roles. Few early-career individuals will fit an RCD role perfectly. Organizations building RCD teams should value curiosity, self-direction, and communication among early-career roles, as these qualities seemed to align with participants’ advancement and growth. Our work also demonstrated that there is a great deal of variance in RCD career paths. PhD’s provided people with meaningful advantages, and for those who came from academia, often led them to the RCD field in the first place. However, PhD’s are not required to have a successful and impactful RCD career. Advancement typically occurred within organizations, regardless of education level. Moves to other organizations were usually lateral in nature, though they opened up further advancement and more personally fulfilling roles. Similarly, mentorship was often very important in earlyand mid-stage careers due to its importance in finding the field, for developing capabilities and qualities that individuals may have lacked, or for its importance to advancement. However, some individuals lacked good mentorship entirely; their career paths were often based on their ability to adapt to shifting organizational needs or environments. Despite these project successes, there is still work to be done. Opportunities include gathering a larger sample size, refining coding procedures, and taking a more quantitative approach to analyze the temporality, directionality, and strength of various factors. This would be valuable as various moderation and mediation hypotheses emerged as a result of our work. Additionally, these would open up possible analysis through graph theory algorithms. 4 Acknowledgements This work has been supported in part by National Science Foundation grants OAC-1620695 and OAC-2100003. Special thanks are owed to the community members who were interviewed for this work, and to our other CaRCC Career Arcs working group members Gladys Andino, Shafaq Chaudhry, and Preston Smith who contributed to the RCD Career Arcs effort. References [1] Lisa Arafune, Dana Brunson, Thomas Hacker, and Preston Smith. Building the research innovation workforce: a workshop to identify new insights and directions to advance the research computing community. Technical report, February 2021. [2] Nicholas Berente, James Howison, Joel Cutcher-Gershenfeld, John L. King, Stephen R. Barley, and John Towns. Professionalization in Cyberinfrastructure. SSRN Scholarly Paper ID 3138592, Social Science Research Network, Rochester, NY, February 2017. [3] Scott Callaghan, Elsa Gonsiorowski, Weronika Filinger, and Ilya Zhukov. Developing a Successful HPC Mentoring Program at the IHPCSS. In Practice and Experience in Advanced Research Computing 2024: Human Powered Computing, pages 1–9, Providence RI USA, July 2024. ACM. [4] CaRCC. Research Computing and Data Professionals - Job Elements and Career Guide, April 2018. [5] Shafaq Chaudhry, Arman Pazouki, Patrick Schmitz, Elizabett Hillery, and Kerk Kee. Understanding Factors that Influence Research Computing and Data Careers. In Practice and Experience in Advanced Research Computing (PEARC ’22), Boston, MA, USA, July 2022. Association for Computing Machinery. [6] Joel Cutcher-Gershenfeld, Torey Battelle, Dana Brunson, Thomas Cheatham, Jacob Fosso Tande, Douglas Jennewein, Julie Ma, Lauren A. Michael, Timothy Middelkoop, Henry Neeman, and Patrick Schmitz. Professionalization of Research Computing and Data: An Expanded Agenda. In Practice and Experience in Advanced Research Computing, pages 129–136, Portland OR USA, July 2023. ACM. [7] Weronika Filinger, Julie Mullen, Jeremy Cohen, Samantha Wittke, and Ann Backhaus. Building HPC Learning Pathways: Understanding our Community. In Practice and Experience in Advanced Research Computing 2024: Human Powered Computing, pages 1–7, Providence RI USA, July 2024. ACM. [8] Ritika Giri, Scott Yockel, Kimberly Grasch, Ashley Stauffer, Stefan Robila, and Chris Reidy. Examining rcd roles and activities through the lens of facings, March 2025. [9] John Goodhue, Julie Ma, Adrian Del Maestro, Sia Najafi, Bruce Segee, Scott Valcourt, and Ralph Zottola. Northeast Cyberteam Program–A Workforce Development Strategy for Research Computing. The Journal of Computational Science Education, 11(1):8–11, January 2020. [10] Thomas Hacker, Preston Smith, Dana Brunson, Lisa Arafune, Thomas Cheatham, and Ewa Deelman. Building the Research Innovation Workforce: Challenges and Recommendations from a Virtual Workshop to Advance the Research Computing Community. In Practice and Experience in Advanced Research Computing, pages 1–7, Boston MA USA, July 2022. ACM. [11] Patrick Schmitz, Scott Yockel, Claire Mizumoto, Thomas Cheatham, and Dana Brunson. Advancing the Workforce That Supports Computationally and Data Intensive Research. Computing in Science Engineering, 23(5):19– 27, September 2021. Conference Name: Computing in Science Engineering. [12] Anselm Strauss and Juliet M. Corbin. Basics of qualitative research: Grounded theory procedures and techniques. Basics of qualitative research: Grounded theory procedures and techniques. Sage Publications, Inc, Thousand Oaks, CA, US, 1990. Pages: 270. 5