How AI EdTech Startups Are Rebuilding African Education from the Ground Up
Africa has a youth problem in the most consequential sense. The continent is home to the world’s fastest-growing youth population, yet its formal education systems have struggled to keep pace. Classrooms are overcrowded. Qualified teachers are unequally distributed. Broadband remains unreliable, especially outside major urban centres.
And yet, a generation of founders is doing something significant. They are not waiting for governments to fix infrastructure before building on top of it. They are building for the infrastructure that exists, and using artificial intelligence to close the gap between what schools can offer and what learners actually need. The results are uneven, the scale modest in many cases, but the pattern is real and growing.
Personalised Learning in an Unequal Classroom
South Africa’s educational inequality is well-documented. The country has some of the continent’s best-resourced private schools and some of its most under-resourced public ones, often within the same city. It was into this gap that Mark Horner and Neels Westhuizen built Siyavula, a Cape Town-based edtech company that began in 2002 as a volunteer project producing open-source mathematics and science textbooks.
What started as a content initiative evolved, over more than a decade, into something more technically sophisticated. Siyavula developed what it calls an “intelligent practice” engine. This cognitive learning platform generates an unlimited number of mathematics and science exercises, individually tailored to each student’s skill level, using machine learning to adapt the difficulty to each learner. The system tracks not just whether students answer correctly, but how they learn, at what pace, at what time of day, and at what point they plateau.
An independent survey found that 97 percent of learners reported the platform had helped their grades improve, and 74 percent said it had improved their quality of life. Those are extraordinary self-reported numbers, and they deserve scrutiny. But the broader picture, a platform with millions of active textbook users and hundreds of thousands of students on its practice engine, suggests genuine traction, not mere novelty.
What is structurally notable about Siyavula is its deliberate decision to keep costs low enough to reach students across income brackets. Through a partnership with MTN, the platform launched a free daily mathematics question service on the Ayoba chat platform, reaching more than 100,000 subscribers across seven territories. The model suggests that AI-driven personalised learning need not be a premium product reserved for those who can afford private schooling.
Closing the Skills Gap, City by City
Move north and west, and a different kind of educational crisis comes into focus: the mismatch between what universities produce and what labour markets demand. Across North and West Africa, graduates are entering job markets without the digital skills that modern employers require. Amine and Yahya Bouhlel, two siblings who had studied and worked in Europe and Silicon Valley, respectively, saw this firsthand when they returned to Tunisia.
GOMYCODE was founded in 2017 and has grown to operate in multiple countries across Africa and the Middle East, with more than 1,000 new students enrolling each month across its 30-plus educational courses. The company’s blended model, combining in-person training centres with an online platform, is designed to solve a problem that purely digital platforms have struggled with in African markets: the difficulty of retaining students who lack reliable internet and who often learn better with some face-to-face instruction.
Amine Bouhlel, the company’s COO, has pointed to a structural reality: around 50 percent of young graduates in the region are unemployed because academic programs fail to deliver the digital skills the labour market needs. GOMYCODE’s response has been to teach those skills directly, in the cities where learners actually are. The startup teaches in twelve local languages and has positioned itself as a regional leader in affordable, localised tech training.
In 2024, GOMYCODE took a further step, launching the GoMyCode School of Technology as an accredited institution offering degrees equivalent to those recognised by European universities, a move that signals ambition beyond short-course certification. The transition from bootcamp to university is rare in African edtech, and it raises questions about how far a skills-training model can scale into formal accreditation without losing the agility that made it effective in the first place.
Building for Bandwidth Scarcity
Perhaps the most persistent structural barrier facing African edtech is connectivity. Even platforms that are technically sophisticated struggle when students spend hours each day in areas with unreliable or prohibitively expensive mobile data. A cohort of younger startups is specifically engineering around this constraint.
In Nigeria, Divine Iloh and Ebuka Osunwoke co-founded SabiScholar, an offline-first secondary school learning platform that takes a direct approach to bandwidth scarcity. The platform’s AI monitors a student’s deadlines and network availability, automatically pre-downloading course materials during off-peak hours or whenever a stable connection is detected, ensuring assignments and videos are available on the device before they are due. The approach has a recently approved patent in Nigeria, and the platform ran a pilot phase with more than 2,900 students sitting national examinations.
Also in Nigeria, Lukman Abimbola founded TalkPDF AI after watching his younger sister struggle with secondary school chemistry, not because she was incapable, but because she was memorising English-language text without understanding it. TalkPDF AI converts textbooks and PDFs into an interactive audio tutor available in English, Yoruba, Hausa, Igbo, and Pidgin. The system includes an “explain-back mode” that asks learners to explain concepts in their own words, refusing to move forward if a student simply recites memorised text. The platform schedules spaced repetition for concepts that learners have not fully grasped, combining cognitive science principles with African-language accessibility.
Both products suggest a broader design philosophy that is emerging in African edtech: solutions built not around ideal conditions, but around the reality that most students face — limited connectivity, multilingual households, and classrooms that are often under-resourced.
The Market Structure Taking Shape
The ecosystem that is forming around AI edtech in Africa is not yet mature, but it is becoming more structured. Africa’s e-learning market is projected to expand from $3.4 billion in 2024 to $7.7 billion by 2033, driven by improving device affordability, declining data costs in several markets, and growing government interest in digital curricula. In November 2025, over 100 leaders from across the education and technology ecosystem convened at an AI for Education Summit in Nairobi, exploring how to embed AI into teacher support, personalised learning, and assessment at scale, a sign that the conversation is moving from pilot programmes to policy.
Institutional support is accelerating. Injini, South Africa’s dedicated edtech accelerator, has distributed over R1 million each to ten ventures focused on assistive technology, rural access, and AI tools, and has published a public directory of more than 70 vetted African edtech startups. Ingressive Capital’s early-stage fund has backed platforms like Klas, a Nigerian educational platform that now connects 5,000 teachers with 300,000 learners across 30 countries.
Still, the structural limits are real. Only 40 percent of primary schools and 50 percent of lower-secondary schools in Africa had internet access as of 2024. That gap does not close with a startup. It requires sustained public investment in infrastructure alongside private innovation in software. The risk, as some observers have noted, is that AI edtech ends up serving the connected and the relatively affluent, reinforcing, rather than reducing, the inequalities it set out to address.
What the Pattern Suggests
The startups gaining the most traction in African AI edtech share certain characteristics. They are, almost without exception, building for constraints rather than ignoring them — designing for low bandwidth, multiple languages, and students who are preparing for specific national examinations rather than abstract skills acquisition. They are also, in several cases, blending AI with human instruction rather than replacing teachers, which reflects both a realistic assessment of what AI currently does well and an understanding of how learning actually works in under-resourced environments.
What is less clear is whether the current pace of innovation can match the scale of the problem. Africa’s youth population will continue to grow. The demand for quality education, education that actually prepares people for the economies they will enter, is not a niche concern. It is a continental imperative.
The founders building in this space are not unaware of that. What they are doing, quietly and with considerable ingenuity, is proving that the problem is not impossible. That may be the most consequential thing they have demonstrated so far.

