Top 62%
dans ce catalogue (1067 pratiques notées)
Les scores se concentrent dans le haut de l’échelle ; la position au sein du catalogue est donc souvent plus parlante que le chiffre seul.
Evidence of impact / measured public value1/3
A rigorous impact evaluation led by the Bordeaux School of Economics, Addis Ababa University and RIDA is measuring effects on farming practices, yields, incomes and cost-effectiveness across 200,000 farmers, but results are not yet published.
Transparency, fairness & accountability2/3
The phased design (a 3-district, 100-agent pilot before scaling) and its methodology are documented by the funder (FID) and implementer (Digital Green), with named independent academic evaluation partners.
Transferability / demonstrated replication2/3
Farmer.Chat already runs in both India and Ethiopia on the same Gooey.AI/GPT-4 platform, in close collaboration with Ethiopia's Ministry of Agriculture, showing demonstrated cross-country replication.
Scalability beyond pilot2/3
Phase 2 already spans 15 districts, roughly 200,000 farmers and 3,000 government development agents, up from 3 districts and 100 agents in the pilot.
Governance, capability & sustainability1/3
Delivery runs through Ethiopia's existing government extension-agent network in partnership with the Ministry of Agriculture, though no independent audit of the AI's advisory outputs themselves has been published yet.
Evidoria. Farmer.Chat en Éthiopie — le pilote de conseil par IA de Digital Green, soutenu par le gouvernement, s'étend à une évaluation d'impact portant sur 200 000 agriculteurs. ID persistant: b6f9dd19-489b-4e6b-b102-b6e0fb6b9051.