Many sectors are guilty of designing tools which already exist, without building on the learnings of what came before, and we’re in danger of doing the same when it comes to AI. Across the sector, agencies are independently building AI chatbots, drafting AI guidance, and commissioning landscape assessments, duplicating efforts for limited gain. Those that aren’t building independently, without their own specific budgets or with limited expertise, are being left behind.

There are some real bright spots though, In Kenya, Jacaranda Health took Meta’s open-weight Llama models and built UlizaLlama, a Swahili language model that powers personalised maternal health advice for mothers using its SMS service. Rather than guarding it, they released the model openly with a developer toolkit, so other African organisations can adapt it to their own domains. In India, the government’s Bhashini platform provides open language AI infrastructure across 22 scheduled languages, built on models from the AI4Bharat research lab and serving millions of daily requests for citizen-facing services. Different continents, different problems, same architecture: open foundations, local adaptation, shared back.

The Frontier Tech Hub (which runs GRTD’s Frontier Tech programme, part of the technology, gender and inclusion pillar of the GRTD portfolio) is launching OASIS with that in mind: a shared, open platform for AI in development. It’s a place to find tools other people have built, discuss what’s working, and build on it rather than starting again. It’s also a place to learn from each other’s mistakes, which in this field still happen often enough to be worth writing down.

Why now
The refreshed Principles for Digital Development say it plainly: sharing, reusing and improving existing initiatives pools the sector’s collective resources and avoids costly duplication. These principles aren’t new. The sector has endorsed them for a decade and, if we’re honest, mostly honoured them. OASIS is our attempt to close that gap, at least a little, by giving practitioners somewhere to actually do the sharing and reusing rather than just agreeing it’s a good idea.

EvalExplorer: evidence in the Ebola response
One of the tools already live on OASIS is EvalExplorer, which uses AI to unlock insights from the thousands of FCDO evaluation reports that currently sit underused. It’s designed to help staff compare interventions and synthesise lessons across whole portfolios, rather than reading report after report.

EvalExplorer was tested under real pressure during the 2026 Ebola outbreak response. The FCDO Evaluation Unit used EvalExplorer to identify which past FCDO investments had been most successful in tackling Ebola, work that would previously have taken weeks. They did it in 3 days. This rapid turnaround meant FCDO could be more confident in the advice developed for ministers and senior leaders, which drew fully on FCDO’s own experience and accumulated learning from previous outbreaks, ahead of international discussions on the outbreak response.

EvalExplorer was also a runner-up at the Government Social Research Awards 2026, recognising Rachel Lineham and Ian Montago from FCDO for “Unlocking Evaluation Evidence with AI.” The judges called it a standout example of how AI can be harnessed to drive better use of evidence in decision-making.

Try it yourself
OASIS is free and open to anyone working in development. Alongside EvalExplorer, it hosts DevExplorer, for AI-assisted analysis across the IATI dataset, and the open code and data behind both tools on GitHub.

Explore the tools at oasis.frontiertechhub.org

You can also join the new OASIS community on LinkedIn: an open space to explore AI tools, open-source code, and learnings we’ve built across the development sector.