Wild Yield

About the team

We are three East African students: a PhD candidate in AI for agriculture working on multifunctional landscapes at CGIAR (Ethiopia), a software engineer and project-management master's student (Ethiopia), and an electrical and computer engineering master's student (machine learning, energy modelling; Rwanda). We build affordable, field-ready technology for smallholder landscapes. Our drive: biodiversity restoration on farms scales when farmers can see it, prove it and earn from it.

Our vision

Across East Africa's smallholder landscapes, biodiversity disappears one field at a time. Hedgerows, woodlots, wetlands and forest edges are cleared because nobody can see their value, prove their recovery or pay for it. The farmer who keeps habitat carries the cost alone, while the benefits (pollination, pest control, soil, water) stay invisible and unrewarded. Conservation data is scarce, expensive and rarely reaches the people who decide how land is used. Our long-term goal: by 2035, farmed landscapes in Ethiopia and Rwanda where biodiversity is measured as routinely as yield; where every farmer and cooperative can see what habitat they have and what it does for them; and where restoring it pays, through better harvests, premium markets and biodiversity finance. We want nature-positive farming to become the economically rational choice for smallholders, starting with one landscape in Ethiopia and designing for the continent.

Our solution

Wild Yield is a landscape intelligence engine for smallholder regions. It closes the loop between planning biodiversity and proving it. 1) SEE. Satellite imagery and machine learning map fields, habitat patches, corridors and pressures (including fuelwood demand) at village scale. 2) PLAN. A multifunctional landscape optimiser recommends where and which practices (hedgerows, riparian buffers, agroforestry, exclosures) restore connectivity and species at the lowest cost to farm income, producing a shared vision map that farmers and local government can act on. 3) PROVE. Cheap, solar-powered acoustic and camera sensors with on-device AI measure birds, pollinators and bats before and after, turning plans into evidence. 4) EARN. Verified outcomes feed cooperatives, certification bodies and biodiversity and carbon finance, so restoration pays. Why now: certification schemes and biodiversity credit markets demand verifiable outcomes; East Africa has almost no affordable monitoring; governments and cooperatives already invest in restoration without measuring biodiversity. What is unique: planning and proof in one system; designed by East Africans for East African realities (offline, cheap, local species); grounded in our research experience in multifunctional landscape modelling. By April 2027: one landscape mapped and planned, a first sensor network deployed, an open dashboard.