Abstract This study evaluates the macroeconomic and cognitive impacts of a dual-track educational transition in sub-Saharan Africa, which integrates advanced EdTech digital public infrastructure (DPI) with culturally-centered pedagogy. Moving away from traditional colonial-era frameworks characterized by rote memorization and foreign-language instruction, this research explores how localized curriculum adaptation combined with machine learning personalization affects student learning outcomes and future labor productivity. Utilizing a quasi-experimental research design across 12 months, the study tracked 600 secondary school students in urban and rural learning environments within West Africa. The empirical findings show that the experimental cohort, exposed to localized STEM learning models delivered via indigenous-language adaptive software, achieved significantly higher cognitive processing and problem-solving scores (\(M = 82.4, SD = 5.2\)) than the control cohort utilizing standard foreign-language materials (\(M = 54.1, SD = 11.8\)). Multiple linear regression modeling confirmed that the combination of digital infrastructure deficits, language barriers, and abstract curricula accounted for \(71.2\%\) of the observed variance in underemployment risks among graduates. The study concludes that an educational transition combining technological fluency with local cultural relevance is an essential requirement for boosting total factor productivity (TFP) and unlocking sustainable 21st-century economic growth across the continent.