Session 1.1 - Understanding career pathways and challenges for RSEs and dRTPs
- Jeremy Cohen, Imperial College London
- James Graham, King’s College London
- Isabella von Holstein, Imperial College London
- Arianna Ciula, King’s College London
This mini-workshop session will focus on the important topic of sustainable careers for RSEs and the wider community of technical professionals in research who are increasingly known as digital Research Technical Professionals (dRTPs). Work to explore or implement new technical roles and career pathways at research organisations often focuses on engagement with institutional leadership. We need buy-in and support from senior leaders, and from HR, to make such changes possible. However, challenges in this space also need buy-in from and engagement with members of the research technical community who will undertake these roles. We need the opportunity to highlight our experiences, the challenges we’ve faced and what sustainable career options look like to us.
The large community of RSEs and other dRTPs that CW brings together, and the collaborative nature of the event, make CW the perfect forum to hear the perspectives of the practitioners themselves around career pathways. What are we struggling with? What have we seen work and what have we seen fail? What is it that we actually want a career structure for our technical roles to look like?
This session will provide an opportunity to develop a picture of community opinion around technical careers. UKRI have supported a number of Digital Research Infrastructure (DRI) projects over the last couple of years and many of these projects have elements that relate to supporting and developing career structures for members of the dRTP community. There are now more opportunities than ever to take the perspectives of our community around careers and feed them into ongoing discussions and activities.
We can’t promise to change the career landscape for dRTPs immediately. However, we can help to connect up a pipeline that takes the views and ideas of dRTPs and ensures that they are fed into the process of developing models for dRTP careers that can most effectively be supported within the complex constraints of research organisations.
The session will begin with an introduction and 4 lightning talks, including talks from people working on technical career pathways (potentially both digital and non-digital). There will then be a chance to anonymously gather some thoughts and statistics from the attendees, followed by group discussions where attendees can discuss their own challenges and aims around roles and career options. Groups will report back at the end of the session and the organisers will produce a short summary report following on from the session.
Session 1.2 - Oops, did we do it right? Lessons from Successful Collaborations
- Eleanor Broadway, EPCC
- Eva, Fernandez Amez, Durham University
Collaborations between initiatives, institutions and individuals are often essential for building sustainable and impactful work, yet they are also where many projects struggle. There are initiatives providing important forms of support; CAKE is supporting new long-term collaborations through funding, while SHAREing is establishing structured learning pathways for Research Technical Professionals (RTP) skill set development. Even so, collaboration is rarely straightforward: it requires intention, planning, and strong foundations beyond funding alone.
This interactive mini-workshop invites participants to reflect collectively on what actually makes collaborations work, and what tends to be missing when they fail. Using a journey-mapping approach, participants will map collaborations they have been part of that were successful, focusing on the full collaboration lifecycle - from early relationship-building through to longer-term collaboration development. Participants will reflect on questions such as: What support was critical at different stages of collaboration? Where did you struggle, and why? What knowledge, communication, or professional skills were missing at the time? This reflection will highlight key priorities, decision points, and the support structures and skills needed to ensure collaborations are successful.
By comparing experiences across participants, the session will identify recurring challenges faced by new or early-stage collaborations and map these against existing training or support resources - highlighting both what already exist and where significant gaps remain.
The aim is not to define a strict single model for successful collaboration, but to build an experience-based resource to help future projects avoid common pitfalls and build stronger, long-lasting collaborations from the very beginning.
Session 1.3 - Escaping the Black Box: An Interactive Workshop on Reproducible AI Research (Modelled after an Escape Room)
- Precious Onyewuchi, Data Science Without Borders, OSPO Now
- Laurah Ondari
Reproducibility is a core requirement for trustworthy research software, and as AI research continually grows, reproducing it can be difficult due to undocumented data practices, opaque experimental workflows, and limited transparency around methodological decisions. These challenges can be particularly serious for early‑career researchers, research software practitioners, and community members transitioning into AI‑enabled research, or just curious to know what’s behind AI systems.
This mini‑workshop introduces reproducible AI research through an interactive, scenario‑based escape room. Participants work collaboratively in small groups to navigate a fictional AI research project that is at risk of becoming irreproducible. Through a sequence of structured challenges (“locks”), participants make key decisions about research questions, data documentation and context, methodological choices, experiment tracking, bias and fairness considerations, and responsible sharing of code and results.
Rather than focusing on technical depth or specific tools, the workshop emphasises reproducibility as a learnable, collaborative research practice embedded throughout the research software lifecycle. Each escape‑room challenge maps to a common point of failure in AI research, encouraging participants to discuss trade‑offs, negotiate constraints, and reflect on the social and ethical dimensions of reproducibility alongside technical considerations.
The session concludes with a facilitated debrief in which participants reflect on decisions made during the exercise and co‑create a lightweight reproducibility checklist that can be adapted to their own research or software projects. Designed for a mixed audience and requiring no prior AI expertise, this workshop supports CW26’s aim of strengthening the research software community by building shared understanding, confidence, and practical capacity for reproducible AI research.
Session 1.4 - Your planet needs you! Scoping a community of Sustainable Computing Ambassadors
- Loïc Lannelongue, University of Cambridge
- Kirsty Pringle, Software Sustainability Institute
- Colin Sauze, National Oceanography Centre
- James Tyrrell, University of Birmingham
- Joe Wallwork, Institute of Computing for Climate Science, University of Cambridge
The research software community's engagement with sustainable computing has accelerated in recent years, evidenced by environmental sessions at Collaboration Workshops and RSEcon, the Green DiSC scheme, and the Green RSE Special Interest Group. Despite this momentum, many individuals championing green computing within their institutions do so informally, without recognition or structural support.
Across research organizations, digital research technical professionals, Research Software Engineers, and researchers act as de facto ambassadors for sustainable computing. They advocate for energy-efficient practices and embed environmental considerations into research workflows. However, this critical work typically falls outside formal job descriptions, leaving champions isolated and their contributions undervalued in career progression.
This interactive mini-workshop establishes foundations for a Sustainable Computing Ambassador network. Through facilitated discussions and role-specific breakout sessions, participants will map sustainable computing activities and co-design template job descriptions adaptable for different professional contexts. These templates will enable institutions to formally recognize this work, allocate time for sustainability activities, and support career advancement.
Drawing on successful models like Data Champions programs, we explore how structured ambassador schemes can amplify individual efforts into systemic change. The Green RSE SIG will refine workshop outputs through continued community engagement, making role templates publicly available as resources for institutions formalizing support for sustainable computing practices.
Session 1.5 - Community Driven Guidelines on Responsible AI in Research Software
- Joe Shingleton, University of Glasgow
- Oscar Seip, University of Manchester/Software Sustainability Institute
- Philipp Boersch-Supan, British Trust for Ornothology
- Sam Harrison, Centre for Ecology and Hydrology
- Loïc Lannelongue, University of Cambridge
The responsible and ethical adoption of Generative AI as a tool for supporting research code production is a growing area of interest amongst research professionals. Existing position papers and blog posts have effectively demonstrated many of the risks and opportunities involved with AI in RSE, as well as the potential impacts on the RSE community. However, a wider community-driven consensus needs to be sought on how these ideas and suggestions are translated into explicit guidelines which are (a) actionable and (b) widely relevant to the needs of the research software community. During this workshop, we will identify a set of practical solutions to mitigate the risks that are most pertinent to the research software community.
Session 1.6 - Exploring community health metrics
- Cassandra Gould van Praag, RCM Cooperative
- Johanna Bayer; Donders Institute for Brain Cognition and Behaviour
- Arielle Bennet Lovell; The Alan Turing Institute, TTW
- Sara Villa; RCM Cooperative, OLS, TTW
- Emma Karoune; RCM Cooperative, The Alan Turing Institute
- Maryblessing Okolie; CHAOSS, OES, OLS, RSECon
The health of a research community is difficult to measure and communicate. Research community managers (RCMs) and those with RCM responsibilities often rely on coarse engagement data from communications channels, but these can lag behind substantive changes in engagement and miss factors relevant to the sustainability of a community. Without broad agreement and adoption of research community health metrics, it is difficult to compare between communities or individual initiatives, and share lessons around successes or challenges in engagement which are backed by scalable data. Existing shared metrics are often based on open source software or commercial communities that can have distinct motivations from research communities, and different normalised modes of working compared to research communities (for example in technology choices), which mean they cannot be directly mapped.
We propose a move towards a common language and shared methodology for measuring research community health. CW26 participants are excellently positioned to contribute to a discussion on both the practical implications of this effort, and an exploration of the ethical concerns relating to the collection of member activity, data which could be perceived as surveillance.
This workshop will bring together RCMs and community members, and contributors to existing metrics projects such as the Community Health Analytics in Open Source Software (CHAOSS) project. We will share experiences of applying community health metrics relevant to academic use cases, and identify common barriers, dependencies or rate limiting factors in their adoption into our workflows. We will work towards identifying next steps in our own practices to enable the communication of more nuanced community health data, and actions we can take collectively to move towards ethical implementation of standardised research community health metrics. We believe this work will support an improved understanding of the factors that contribute to a healthy research community, as relevant to both community managers and community funders.
Session 1.7a - Narrating FAIR: recognising and rewarding FAIR practices and the use of narrative CVs
- Samantha Ahern, Society of RSE
FAIR is used to talk about three things: data (or any digital object), metadata (information about that digital object), and infrastructure, although data is the most commonly used term. Over £2trillion is spent globally each year on research and development that produces a range of digital objects which are useful for further and future research. However, these are often difficult to find and reuse by people to bring continual benefits to society.
Only 28% of researchers are familiar with the principles ( 2021 State of Open Data report by Digital Science), with fewer enacting those principles. A recognised barrier to enacting FAIR (Findable, Accessible, Interoperable and Reproducible) practices in researchers working in practices is a lack of recognition of both the researchers' efforts and the public good that is making digital objects available in a FAIR way such as:
- maximising the potential of research outputs by allowing researchers to reuse and repurpose outputs in novel and innovative ways.
- greater accessibility and transparency of the research process.
- enabling replication and verification of research findings which can also lead to more collaboration.
Session 1.7b - Colandr: Demonstrated Impact in Evidence Synthesis, Open for Community Innovation
- Gavin Fulmer, PhD, Senior Manager for Research, DataKind
Systematic reviews are critical for building robust scientific knowledge, but manually screening hundreds or thousands of citations is slow, expensive, and difficult to reproduce. Colandr (www.colandrcommunity.com) is an open-source, computer-assisted platform that has been used for over a decade to streamline this process using active machine learning, while retaining human oversight and collaborative workflows. By combining AI support with peer collaboration, Colandr ensures transparent, rigorous, and reproducible evidence synthesis.
Originally developed for conservation science, Colandr has demonstrated impact across health, education, social development, and environmental management. It is particularly valuable for teams in low- and middle-income countries where commercial tools may be cost-prohibitive. With a decade of real-world application, Colandr has proven its ability to reduce screening time, support reproducibility, and increase accessibility for diverse research communities.
DataKind is ready to take Colandr to the next level – and we're looking for partners to help shape and extend it into the future.
In this session, we will demonstrate Colandr 2.0: a fully rebuilt, stable, and extensible platform designed for community contribution and scalable workflows. We will show how Colandr supports collaboration, reduces researcher burden, and thoughtfully integrates AI without compromising transparency and human judgement. Participants will leave with a clear understanding of how to adopt, contribute to, or advocate for Colandr – and how to help shape its ongoing evolution as a widely used, open, and impactful research tool.
This session is ideal for researchers, software engineers, funders, and anyone involved in systematic reviews or evidence synthesis who wants a practical, proven, and community-oriented tool for their workflow.
Session 2.1a - Explore the assistive technology built into your computer
- Eli Chadwick. University of Manchester
- Alexander Hambley, University of Manchester
- Jim, O'Donnell University of Oxford
- Pao Corrales, 21st Century Weather
- Jyoti Bhogal, RSE Asia Association
Are you interested in how disabled people adapt their devices for their needs? And do you want to make sure your software supports them?
Almost all operating systems come with built-in accessibility features and assistive tools. These facilitate device access for people with all kinds of disabilities - including vision, hearing, mobility, and cognitive disabilities. As software creators, it's essential to be aware of how these different tools change the way someone interacts with the software we build. And a great way to learn is to try them!
In this workshop you will learn how to enable and use a variety of built-in tools and settings on your computer designed to support people with disabilities. Anyone is welcome to attend - whether you are completely new to learning about disability and assistive tech, or you are a regular user of one type of assistive tech who is curious about tools designed for other needs.
You will leave with a better knowledge of the built-in accessibility tools and features that are available on the operating system you use. You'll also be able to use some of those tools to perform basic software testing. You may discover features that you find useful even if you do not have a disability!
Operating systems covered: Windows, MacOS, Linux (Ubuntu).
Session 2.1b - Poetic computing 101 (and other approaches to software)
- Anne Steele, Missing Maps
How can creativity hack computation? Does adversarial poetry reveal something new about large language models? Can live coding change our relationship to the browser? What can permacomputing teach us about (software) sustainability?
This demo session is a brief introduction to the world of creative computing, a world that exists in parallel to traditional research software, but may offer alternative perspectives to notions of sustainability, creativity, and care.
Participants will be introduced to an ecosystem of existing work in the field, and experiment with simple tools and prompts to observe how computational systems shape meaning, behaviour, and sustainability. No prior experience in programming (or poetry for that matter) is required – only curiosity and a willingness to experiment.
Session 2.1c - The Role of Research Software in Sustainable Computing
- Anne Steele, Missing Maps
The environmental footprint of high-performance computing and cloud computing infrastructures is high both at the time of constructing computing equipment and when in operation. An immediate impact shows itself in high capital and operational costs. This workshop explores the role of research software engineering in the sustainability of computing through short talks and panel discussion. It is organised in the context of the Kelvin Living Lab: Towards Net Zero High-Performance Computing (EP/Z531054/1).
Session 2.2 - Building Stronger Communities: Insights From Multiple Perspectives
- Maria Doyle, University of Limerick / Bioconductor
- Kevin Rue-Albrecht, University of Oxford
- Nick Cooley, University of Limerick
- Alan O’Callaghan, University of Edinburgh
- Ella Kaye, University of Birmingham
- Heather Turner, University of Birmingham
- Sara Villa, RCM Cooperative, OLS, TTW
- Cassandra Gould Van Praag, RCM Cooperative
- Emma Karoune, RCM Cooperative
- Malvika Sharan, RCM Cooperative
- Johanna Bayer, Donders Institute
This mini-workshop explores what genuinely strengthens a research community by bringing together multiple perspectives from the Bioconductor and R ecosystems, as well as other open research communities - including community management, developer engagement, computational biology and SSI Fellow perspectives, contributor viewpoints, and broader ecosystem insights. Each speaker will deliver a short lightning talk reflecting on one effective practice they have seen work in real communities and one barrier that limits participation or sustainability. We will also include a brief cross‑community reflection highlighting how similar practices and challenges appear in other scientific communities, and how they compare with more mature ecosystems such as Bioconductor. Following the talks, participants will join structured, facilitated breakout discussions focused on practical challenges and solutions for strengthening research communities across diverse ecosystems. The workshop emphasises people, collaboration, roles, and cross-community learning, directly supporting the CW26 theme of 'Strengthening the Research Software Community.' Expected outcomes include shared community principles, renewed awareness of effective practices, stronger cross-role understanding, and a post-event write-up capturing key insights.
Session 2.3 - Analysing AI guidelines to critically enquiry AI adoption in software development (in-person only)
- Carlos Cámara-Menoyo, University of Warwick
Please note that the organiser will be providing some policies from HE organisations. However, contributions from other participants will be highly appreciated to enrich the discussion. Those wanting to collaborate are kindly invited to send documents or links to publicly available policies by email to carlos.camara@warwick.ac.uk.
AI is profoundly transforming many facets of our lives, one of them being how we think about and develop software. Specialised LLMs, IDE integrations, a myriad of services and platforms, as well as the emergence of new development methods (e.g. vive coding) all of which promise speed and efficiency, have led to a massive adoption in corporate environments. This shift has also permeated to HE organisations who are also embracing AI to develop research software (as well as for teaching and researching) even though they collide with their core values and principles such as scientific knowledge, academic integrity, inclusion and diversity, collaboration or committing to producing a positive impact beyond academia, usually related with sustainability or social justice.
This workshop will be the kick-off of my SSI Fellowship aimed at Creating CriticAI, a community of practice to critically enquiry about AI’s adoption in HE and influence in the decision-making processes to ensure that AI usage is aligned with academia’s ethos. In the first half of this session, participants will be analysing universities’ guidelines to produce a collaborative mapping of AI’s adoption across HE. This will inform a discussion on the second half of the session, where we will collaboratively be exploring the challenges, tensions and contradictions that AI is introducing in our jobs and workplaces.
Those attending to the workshop are expected to meet and engage with a group like-minded software developers who want to and gain a renewed critical awareness of AI adoption that probably (and hopefully) can lead to form a community of practice and collective action.
Session 2.4 - How should the DIRECT Framework adapt to serve diverse dRTP communities of practice?
- David Horsfall, Newcastle University
- Phil Reed, The University of Manchester
- Aleks Nenadic, The University of Manchester
- Mike Simpson, Newcastle University
- Kirsty Pringle, EPCC
- Adrian D’Alessandro, Imperial College London
- Sara Villa, OLS/ TTW
- Emma Hogan, Met Office
- Samantha Ahern, UCL
The DIRECT Framework is a tool to identify competencies, define development pathways and discover resources to build the right skills and progress your career in research software. It is a general framework designed to capture the wide spectrum of skills used across digital research roles - technical competencies (such as programming, software design, data management, research infrastructure), and professional competencies (such as teamwork, communication, project management, leadership, and community engagement).
The framework is open and evolving. Over the past three years, the DIRECT Framework has been shared and refined by feedback from the RSE and other digital research communities through contributions at conferences hosted by the Society of RSE, SSI and others, with further support provided through Network+ and UKRI funding.
The RSE and digital Research Technical Professional (dRTP) community has demonstrated the power of strong, values-driven engagement, with successful initiatives emerging around shared priorities such as sustainable and green research, mental health, teaching and training, and community management. These initiatives thrive when communities can shape and adapt shared resources to meet their own needs. Reflecting this, the DIRECT team has been approached by several communities interested in incorporating, collaborating on, or adopting aspects of the framework within their own activities.
This workshop will bring together members of the RSE and wider dRTP community to co-design sustainable ways for the DIRECT Framework to engage with, and be shaped by, diverse communities of practice. Participants will work with the DIRECT team and partner initiatives, including ConveRSE (mental health in RSE, led by Mike Simpson), the Greening Digital Research project (led by Weronika Filinger with Jeremy Cohen, Martin Jukes and Kirsty Pringle) and the RSE Teaching and Training SIG (led by Samantha Ahern, Liam Berrisford and Aleksandra Nenadic) to identify concrete requirements for adopting, extending, or contributing to DIRECT. The workshop will explore how the framework should evolve to better support community-led priorities such as mental health, EDIA, teaching and training, and environmentally sustainable digital research. The outcome will be a shared, practical model for collaboration: clearly defining how individuals and groups can contribute to the DIRECT Framework, how those contributions are reviewed and sustained, and how the framework can remain responsive to emerging needs across the dRTP ecosystem.
Join the workshop to participate with DIRECT, to learn how you might apply the skills from the framework in your domain, and to have your say in how DIRECT could better serve your community.
Session 2.5 - Small (sample) tests, big impact: usability evaluation for research software
- Christina Bremer, University of Cambridge
- Loïc Lannelongue, University of Cambridge
- Laurent Gil, University of Cambridge
- Anica Araneta, University of Cambridge
- Jyoti Satnam Singh Bhogal, RSE Asia Association
High-quality research software is not only technically robust, it also helps researchers work more efficiently, avoid errors, and focus on their work. To enable this, we need thoughtful, usable interfaces in addition to clean, well-written code. This interactive mini-workshop introduces software practitioners to practical ways of evaluating the usability of their research software. They will learn how to identify usability problems early, gather meaningful feedback from users, and use that feedback to improve the design of their software’s interface.
During the first half of the mini-workshop, we will cover usability principles and hands-on evaluation methods; this includes heuristic evaluation, task-based usability testing, and think-aloud studies. Examples from existing research tools will illustrate the importance of usability as a key consideration in software development and how small changes can make a big difference to users. During the second half of the mini-workshop, participants are encouraged to bring their own software project. With guidance from the facilitators, they can apply what they have learned to their own work, e.g., create a usability evaluation plan, test an interface, or uncover issues that they can address immediately, depending on what is most useful to them. This way, they leave with both background knowledge and practical insights that they can also share with their teams back home.
The mini-workshop will be a 60-minute, hybrid event. Everyone with an interest in the user experience (UX) of research software is welcome to attend. If participants do not wish to bring their own project, they can either choose to work collaboratively on another participant’s project or evaluate an interface that is provided by the organisers.
Session 2.6 - RSE across languages: how do we strengthen collaboration and community?
- Sofía Miñano, University College London/Sainsbury Wellcome Centre
- Salma Thalji, Technical University of Munich/Max Planck Institute for Biological Cybernetics
Research Software Engineering was first formally defined in the UK, but is increasingly supported by communities across several countries and languages (see https://society-rse.org/international-rse-organisations/). However, English remains the de facto interface through which roles, practices, and careers are recognised and shared. This can limit the sharing of practices and resources, effective collaboration and ultimately weaken the research software ecosystem as a whole. At the same time, many RSEs already are multilingual, suggesting untapped potential for strengthening cross-language visibility and collaboration within the global RSE community.
In this 60-minute session, we aim to explore the current status of multilingualism in the research software ecosystem. Specifically, we will examine how the RSE role is articulated across different linguistic and cultural contexts, and gather perspectives on where language supports or hinders inclusion, collaboration, and sustainability within the research software community. We aim to gather feedback from attendees on how RSE communities working in English and non-English contexts can build sustainable, reciprocal collaboration and knowledge exchange.
The session will be exploratory and discussion-led. After a short framing using examples of how the RSE role is described and understood in different national and linguistic contexts, participants will be invited to share experiences from their own settings and reflect on how language choices affect collaboration, recognition, and career development. We will also gather practical ideas and experiences on how RSE communities might better support multilingual participation, including community practices, documentation approaches, and the use of translation or internationalisation tools where relevant, while reflecting on their limitations and trade-offs. The discussion will aim to identify a small number of recurring friction points and potential leverage areas, which may inform follow-up activities such as an unconference session or a CW Hack Day pitch.
The goal of the session is not to propose a single solution, but to surface common challenges and identify opportunities for strengthening collaboration and community across languages within the global research software ecosystem.
Session 2.7 - The business canvas - a stepping stone to software sustainability
- Alexandra Simperler, Simperler Consulting
You have developed this great piece of research software and you have this feeling it could go beyond your university department, should go places and be used by many people. Maybe you want it to become a widely used open source tool. Maybe you are considering a commercial path. Either way, thinking about sustainability and long term viability is essential.
A business canvas is a good first step to think about the viability and future of your software; what is the value of what you are doing and who could benefit from it? It also prompts you to consider practical questions that often get overlooked: Which channels will you use to promote your software? What hidden costs might appear as your user base grows? Working through the canvas does not force you into a commercial mindset—it simply gives you a clearer picture of how your software can thrive, whether it is free and open source or commercial.
Finally, we will delve into the art of the elevator pitch – a concise way to spark interest and encourage stakeholders to dive deeper into your full plan.
