MichaelKamukulu

Michael Kamukulu started his career with a simple task: go into schools and find out what is actually happening. Not what the reports say. Not what the program design intended. What children are actually learning, and what is getting in the way.

That question has not changed since 2012. Everything else has scaled around it.

He began at EDI Global as a research assistant, sitting inside Tanzania's public education data at ground level — present in classrooms, in district offices, in the spaces between what policy says and what teachers do on a Tuesday afternoon. Over six years, he developed into one of Tanzania's most capable education data and program design professionals, not through a neat career path but through proximity to evidence that most people never get close enough to take seriously.

In 2018, he joined Twaweza East Africa to lead KiuFunza — a program built on one straightforward belief: that what matters is not whether children attend school, but whether they learn. That distinction sounds obvious. It is not how most systems are designed. Making it operational at scale required more than a new incentive structure for teachers. It required building the data infrastructure to know whether learning was actually happening: live dashboards that surfaced problems in weeks not years, rapid assessment tools built for real field conditions, automated systems that kept accountability from collapsing under its own administrative weight. KiuFunza reached over 330,000 children. The cost: $7 per child per year. The evidence was rigorous enough to be published in the Quarterly Journal of Economics — placing it among the most credible learning interventions ever tested in Africa.

Across 895+ public primary schools and more than a decade of field work, one conclusion has become difficult to argue with: the architecture of the system determines what people can become. A child surrounded by a system that produces learning becomes a learner. A professional surrounded by a system that produces growth becomes more productive. The failure in both cases is never the person. It is always the design.

In 2025, Michael co-founded LearnImpact as an independent Tanzanian institution and became Executive Director in February 2026. LearnImpact's mandate is precise: work directly with governments to design learning systems that they can own, sustain, and scale — aligned with Tanzania's Vision 2050, the 3Rs framework, and the real demands of an economy where foundational learning and lifelong employability are not separate policy conversations but two ends of the same urgent problem.

The AI dimension of this work is not cosmetic. The systems that will govern how people learn — at seven years old and at thirty-seven — are being designed right now. The decisions being made today about how AI integrates with public learning infrastructure, teacher development, and national assessment will shape employability and productivity across Africa for a generation. Michael's position is that those decisions cannot be made for African governments by institutions that have never been accountable to them. LearnImpact is building the alternative, from the evidence up, in partnership with the governments that will have to live with the results.

Beyond institutional work, he founded the LENZI Impact Network — a learning and professional development platform reaching over a million people across Sub-Saharan Africa. The premise is the same as KiuFunza: when people have the right systems around them, they perform at a level they could not reach alone. LENZI operates at the individual end of the learning arc — helping professionals build the skills, visibility, and capability to stay productive in an economy being reshaped by technology. Both organisations are expressions of the same argument: learning is not something that happens in school and stops. It is the mechanism by which people become productive, and by which economies grow. Systems designed with that understanding produce fundamentally different outcomes than systems that don't.

Fix the learning. Fix the system. The rest follows.