Keeping Research Software Relevant for Tomorrow

Transcript

1. Paths for Impactful Software Tania Allard, PhD | she/her Keeping Research Software Relevant for Tomorrow RSECon 2025
2. Paths for Impactful Software Tania Allard, PhD | she/her Keeping Research Software Relevant for Tomorrow RSECon 2025
3. Tania Allard CEO, Impact - Quansight September 2025 RSECon
4. What do you do at Quansight? September 2025 RSECon
5. Work towards a sustainable open source ecosystem where critical software and infrastructure are robust, secure, accessible, and a collective responsibility. September 2025 RSECon
6. Sustainable development is: Development that meets the needs of the present without compromising the ability of future generations to meet their own needs. Brundtland Report, 1987 Users, community members, contributors, maintainers September 2025 RSECon
7. It is a very tricky balancing act Photo by Jeppe H. Jensen on Unsplash More than 90% of companies rely on open source software (OSS). The OSS ecosystem is not as healthy as it should be. September 2025 RSECon
8. It is a very tricky balancing act Photo by Jeppe H. Jensen on Unsplash Companies are using OSS but not supporting it enough More than 90% of companies rely on open source software (OSS). The OSS ecosystem is not as healthy as it should be. September 2025 RSECon
9. What does this have to do with research software? September 2025 RSECon
10. RSEs and Researchers are a critical part of our ecosystem https://arxiv.org/abs/1907.10121 September 2025 RSECon
11. RSEs and Researchers are a critical part of our ecosystem September 2025 RSECon
12. Incentives are still misplaced in academic and research environments. Grants/funding favour short-term innovation vs. long-term maintenance It is a very tricky balancing act September 2025 RSECon
13. Incentives are still misplaced in academic, research, and corporate environments. How OSS/research software development & maintenance are measured, incentivised, recognised, and funded matters It is a very tricky balancing act September 2025 RSECon
14. Incentives are still misplaced in academic, research, and corporate environments. It is a very tricky balancing act September 2025 RSECon Look at the formalisation of RSEs as a critical role OSPOs are still rare How OSS/research software development & maintenance are measured, incentivised, recognised, and funded matters Academic/research institutions & funders have not been great at this
15. Where is the spotlight? Python is a great choice due to its versatility as a general programming language and its thriving ecosystem https://lp.jetbrains.com/python-developers-survey-2024/#purposes-for-using-python https://github.blog/news-insights/octoverse/octoverse-2024/ September 2025 RSECon
16. Where is the spotlight? Are we still going to need developers/RSEs? How far will big tech reach? Will we become dependent on API providers? Will AI replace all of us? Will it take over open source software? How do we keep data private/ open? How can AI enhance research? How do I detect slop in research software? September 2025 RSECon
17. Where is the spotlight? Following the money 💰 Are we still going to need developers/RSEs? How far will big tech reach? Will we become dependent on API providers? Will AI replace all of us? Will it take over open source software? How do we keep data private/ open? How can AI enhance research? How do I detect slop in research software? September 2025 RSECon
18. Don’t confuse raising the floor with raising the ceiling. September 2025 RSECon
19. Ceiling Floor High value High adoption AI assistant app adds summarisation capability for research papers +0.5% user growth 📓 💸 Hotel booking platform Improves recommendation system +0.5% click-through rate, +1M revenue 👩💻 Local business sets up website without requiring help from a developer +10% customers and revenue 🔬 ML-driven research accelerates drug candidates discovery process -15% costs (?), opportunities for untreated diseases September 2025 RSECon h.t. Ines Montani - explosion.ai 💥
20. Ceiling Floor High value High adoption AI assistant app adds summarisation capability for research papers +0.5% user growth 📓 💸 Hotel booking platform Improves recommendation system +0.5% click-through rate, +1M revenue 👩💻 Local business sets up website without requiring help from a developer +10% customers and revenue 🔬 ML-driven research accelerates drug candidates discovery process -15% costs (?), opportunities for untreated diseases September 2025 RSECon h.t. Ines Montani - explosion.ai 💥 Low transformation/return https://fortune.com/2025/08/18/mit-report-95-percent- generative-ai-pilots-at-companies-failing-cfo/
21. Ceiling Floor foundational frameworks AI features September 2025 RSECon h.t. Ines Montani - explosion.ai 💥
22. Floor foundational frameworks AI features open research & research software September 2025 RSECon h.t. Ines Montani - explosion.ai 💥 Ceiling Adoption Development & maintenance
23. High-value use cases are worth development and maintenance effort September 2025 RSECon
24. The development process lets you take control (+ focus on high-value stuff, opportunities for innovation and creativity that move the impact needle) Will AI replace all of us? No September 2025 RSECon Maintenance translates to long- term impact and relevance (+ usability, accessibility, security, reusability, extensibility, interoperability, sovereignty, persistence) This is where research software shines Move from application to framework or platform
25. The development process lets you take control (+ focus on high-value stuff, opportunities for innovation and creativity that move the impact needle) Maintenance translates to long- term impact and relevance (+ usability, accessibility, security, reusability, extensibility, interoperability, sovereignty, persistence) It’s a balancing act! September 2025 RSECon
26. Beyond code - maintenance is all encompassing Goals and metrics People Infrastructure Technology Culture Processes Systemic change: we need funding and incentives that support and prioritise sustainability (high value efforts) September 2025 RSECon
27. Openness as a core principle In the age of AI and mistrust, openness is our greatest asset Transparency builds trust It enables and fosters collaboration Ensures long-term access and availability It is a strategy for resilience and impact Raises the ceiling September 2025 RSECon Open… software, science, research, protocols, data
28. Invest in community, people, and governance Balance innovation with dedicated maintenance Focus on raising the ceiling vs. raising the floor 👩💻 ⚖ 💸 Champion openness as a core principle 🤝 Continue to advocate and work for systemic/ culture change 📣 September 2025 RSECon
29. Open source and top-notch research do not run on secret sauce. Knowledge gets shared and built upon. This is how we keep research software relevant. September 2025 RSECon
30. Thank you! quansight.com tallard@quansight.com September 2025 RSECon