EIDU — AI-Adaptive Pre-Primary Learning in Kenya's Public Schools (Peer-Reviewed RCT)

Kenya's public pre-primary classrooms are typically large, under-resourced and taught without individualised feedback, making it hard for teachers to track which 4-to-6-year-olds are falling behind in early literacy and numeracy. EIDU, a Berlin-headquartered non-profit (gGmbH) founded in 2015 by Bernd Roggendorf after he moved into Nairobi's Kibera settlement, built a low-cost digital personalised learning (DPL) tool to address this directly inside government schools rather than as a parallel private offering.

The intervention combines two components: training teachers in structured pedagogy, and an AI-adaptive smartphone application — built on digital content from Onebillion — that gives each child individually sequenced literacy and numeracy exercises based on their demonstrated mastery, with teachers receiving data on class and pupil progress. Delivery cost is reported at roughly $11 per child per year.

Between October 2022 and October 2023, EdTech Hub, researchers from the University of Manchester and Women Educational Researchers of Kenya (WERK) ran a stratified, two-arm cluster-randomised controlled trial across 291 government pre-primary classrooms (1,955 learners aged 4–6), comparing EIDU classrooms to controls at baseline, midline and endline. The trial found treatment effects of roughly 0.425 SD at midline and 0.534 SD at endline on combined literacy/numeracy outcomes — a large effect size by the standards of international education RCTs. The full results were published in a peer-reviewed journal in early 2026. An earlier, separate evaluation of the structured-pedagogy teacher-training component alone found a smaller 0.25 SD gain in school-readiness scores. As of recent reporting, EIDU serves roughly 94,000 children in Kenya, with a stated ambition to reach 2 million pre-primary learners by 2026, and the organisation has begun open-sourcing parts of its platform.

The published trial does not report a gender-disaggregated breakdown of results, so the equity claim rests on EIDU's low per-child cost and its placement inside the public system that serves low-income families broadly, rather than on documented girls-specific outcomes.

Read the full analysis: https://eidu.com/kenya/