Deadline: October 5, 2026
Applications are open for the Digital Science Catalyst Grant 2026. The Digital Science Catalyst Grant is an international initiative to support innovation in new software tools and technologies to advance research and create meaningful change. The program supports and invests in early-stage ideas in the novel use of technology with an award of up to £25,000 for the most promising ideas that aid science and further its impact on society.
AI is starting to do research, not just describe it. The next breakthroughs will come from agentic workflows, autonomous and multi-step, that researchers, institutions and funders can actually rely on. For 2026, they are backing the early-stage tools, agents and platforms making that real: AI that acts on research, embedded in the workflows it serves, with provenance, governance and accountability built in.
Benefits
- The award is up to £25,000, but the grant’s real value is the conversations it starts. Our team comes from research and research software, with decades of experience between us. Most applicants do not win an award, but they can come away with advice, introductions and a sharper idea. We invest in the community because we come from it.
Eligibility
They want early-stage startups, research projects and moonshot thinkers building agentic tools that:
- Operate across multiple steps, planning, executing and reviewing, rather than answering a single prompt;
- Embed in the workflows researchers, institutions and publishers already use;
- Carry audit trails, governance controls and provenance, so the people who rely on them can trust them;
- Tackle a real problem across the research lifecycle: writing, discovery, evidence synthesis, peer review, funding decisions or research integrity.
Application
How to test your idea before applying:
- Name the research decision your agent affects, and who makes it today (who, when, how often).
- Say where it sits in the research lifecycle: writing, discovery, evidence synthesis, peer review, funding decisions or research integrity. If it isn’t one of these, it probably isn’t a fit this year.
- Name the tool or system it has to live inside, and who already uses that tool every day. They are more interested in something embedded in real work than a standalone demo.
- Check the work genuinely needs more than one step. If a single well-written prompt does the job, it isn’t an agentic workflow.
- Test where it breaks. Give it a task it should refuse, flag or escalate, and record what actually happened.
- Name your outcome metric, for example reviewer hours saved, time to funding decision, or duplicate work avoided.
For more information, visit Digital Science Catalyst Grant.
