Case studyAgriculture & AI
In the CGIAR AI Hub's product team, I help turn science program ideas into product stories that partners can build, and strengthen the Human Centric Design practice behind them, across research centres.
CGIAR is one global research partnership. Its accelerators feed its science programs, and the science programs take proven ideas to impact. The AI Hub asks how AI can help that science reach people faster.
The hub is funded by the UAE through its International Affairs Office (IAO), and it builds on the UAE's AI and agricultural innovation ecosystem: MBZUAI for advanced AI research, Inception for prototyping and development, and AIM for Scale for taking solutions to the people who use them.
I'm part of the AI Hub's product team as a service designer, and our role sits in scoping. Together we run the research and stakeholder mapping that scoping needs, facilitate the workshops where ideas get shaped, write the product story at the heart of each pitch, and prepare the handover. Prototyping and development are carried out by Inception or other implementation partners.
Science programs spot real needs on the ground; the hub scopes them, shortlists the strongest, shapes them into one product pitch, and hands them to partners to build. It has already produced live products.
An idea, grown
Science programs identify real needs on the ground, where AI could genuinely help.
The hub checks feasibility and data readiness before anything moves forward.
A competitive process picks the strongest opportunities, not the loudest ask.
The hub and the science programs shape each opportunity into one cohesive, product-based pitch.
Once the International Affairs Office signs it off, the idea is handed to Inception or other implementation partners, who prototype, build, test and scale it.
A working product. This process has already produced live products.
LivestockAI is one of the ideas we took through to a product story. It's now in development.
My part, with the product team: the product story and a farmer-question mock-up.
Feed is one of the biggest costs of producing milk or meat, and balancing a ration is too complex for most farmers. Good tools exist, but they're scattered and built for experts.
A trusted, scalable system that delivers clear livestock feed guidance, to improve productivity and farm resilience in low-income regions.
We started from what already existed: decades of CGIAR livestock and feed research, and the tools built on it over the years, like OFA and RationSmart. We mapped them, then talked to the channel partners who already reach farmers and the implementing partners who would build and run the product. From that, we wrote the product story for LivestockAI: clear, local feed guidance for livestock keepers, extension workers, feed producers and policymakers, built on the tools they already trust.
To make the product story tangible, I mocked up the moment that matters most: a farmer sharing a photo of an animal and the feed on hand, and getting a clear feeding plan back.
The Human Centric Design Champions programme strengthens design practice across CGIAR: each centre has one named champion who connects the central team's work to local projects. Connectors and advocates, not trainers.
My part, with colleagues: contributing to the tools, running the deep-dive interviews with each champion, analysing the results, and facilitating the workshop.
A short survey each champion circulated in their own centre, about one real product:
A longer conversation with each champion, about what happened in practice, not what was intended:
A practice that depends on people more than on systems, with needs that differ from centre to centre.
The findings feed the agenda of the champions' in-person workshop in Morocco. I'm part of the team facilitating it, including the deep-dive gallery session. The days move from experience to network:
