Digital Gender – Inclusive Tools for Evidence Use

In the high-pressure world of global health and gender equality, policymakers often find themselves at a frustrating crossroads. On one hand, they can rely on the gold-standard systematic evidence provided by institutions like ACRES. It’s a process they trust deeply, but it is manual and time-consuming, often taking weeks to deliver; time that a rapidly closing policy window simply doesn’t allow.

On the other hand, the lure of instant answers from public AI tools like ChatGPT is strong, yet these platforms are fraught with hallucinations and outdated data that make them risky for high-stakes decision-making.

The Digital Gender Inclusive Tools for Evidence Use (DG-Tools) project was born from a realisation that the problem isn’t a lack of evidence, but a broken supply chain. To fix it, we combined the rigorous standards of African research with the processing power of artificial intelligence. We set out to build digital tools capable of doing in minutes what traditionally took a full research team weeks to accomplish.

This journey was a continental response that brought together a unique blend of expertise. Supported by the Bill & Melinda Gates Foundation, the project united ACRES’ expertise in evidence synthesis and rapid response services, the technical prowess in building artificial intelligence of Sunbird AI in Uganda, the public health insights of the Ethiopian Public Health Institute (EPHI) and the evidence-use experience of the Pan African Collective for Evidence (PACE) in South Africa. This partnership ensured that the tools were co-created by evidence experts, technologists and end-users; the very people who understand the complexities of African policy contexts.

The results of this collaboration have moved beyond theory into tangible assets that are already reshaping the EIDM landscape. We developed The Living EIDM ToolMap, an open-access resource that curates and maps the digital software ecosystem to help researchers avoid duplicating efforts, save time, and accelerate informed tool selection. More importantly, we proved the technical feasibility of autonomous synthesis through the ACRES DataExtractor, Dr-RES and the cRhona platform. These tools use Retrieval-Augmented Generation (RAG) to ensure every AI-generated answer is grounded in verifiable, scholarly databases, directly solving the trust gap inherent in public AI.

Ultimately, DG-Tools has redefined what rapid means in the context of evidence synthesis. By moving timelines from weeks to near-instantaneous responses, we are ensuring that the most vulnerable; and the gender-inclusive data often overlooked, are represented in policy decisions.

This project stands as a validation that AI, when harnessed responsibly, can be a powerful equalizer in the quest for timely, effective, and equitable governance.