South African EdTech incubator Injini has launched an eight-month AI for Education Venture Builder and Incubation Programme, backed by philanthropic funder Coefficient Giving, aimed at taking early-stage teams from a raw idea to a tested AI-powered education product. Twelve teams from across Southern Africa will be selected to build, with the five strongest ventures advancing to a final incubation phase and each receiving $25,000 in equity-free seed funding to keep developing their product afterward.
Who Can Actually Apply
The programme is open to pre-formed teams of at least two people, though Injini says teams of three or more are preferred, based in South Africa, Botswana, Eswatini, Lesotho, Malawi, Mozambique, Namibia, Zambia, or Zimbabwe. Each team needs to collectively cover three roles: a builder who can actually bring the product to life, a domain expert or context lead with real understanding of the education problem being solved, and a commercial or operations lead who can think through the business side. Applicants need two to three years of relevant work or lived experience and must be at the ideation stage, meaning a concept that’s ready to be developed but hasn’t launched yet. Applications close at midnight on September 27, though Injini is reviewing submissions on a rolling basis, with successful teams announced on October 21.
Why Injini Says This Needs to Exist
Injini Executive Director Krista Davidson framed the programme around a specific, uncomfortable gap: Africa is significantly underrepresented in the data used to train AI tools, and without deliberately building homegrown solutions now, the continent risks ending up with EdTech that simply doesn’t work for African classrooms. That’s a different problem than lack of ambition or lack of good ideas. Davidson has described spending years talking to people across the region who understand education problems intimately but lack the team, capital, or clear path to actually build something, which is precisely the gap this venture builder is designed to close. The specific challenges the programme is built around are concrete rather than abstract: severe teacher shortages, gaps in foundational literacy and numeracy, linguistic diversity across the region’s classrooms, and ongoing connectivity and device affordability constraints that shape what any realistic solution can actually look like on the ground.
The Five Areas Teams Can Build In
Injini has organised the programme around five thematic focus areas: foundational learning, teacher capacity and classroom support, learner pathways and futures, inclusion and access, and ecosystem and engagement. That spread covers everything from direct classroom tools to the broader systems and support structures around teaching and learning, giving applicants real room to interpret what an AI-powered education solution actually means for their specific context.
How the Eight Months Actually Break Down
The programme runs in three stages. Selection and onboarding takes place through September and October, opening with an in-person orientation bootcamp in Johannesburg. From November 2026 through March 2027, all 12 selected teams work through problem validation, technical design, building a functional AI-powered minimum viable product, and early user testing, with travel and accommodation for up to three members per team covered for mandatory in-person engagements. From April through June 2027, the five ventures that make it through incubation get additional support to refine their product, run further testing, measure real educational impact, and build out go-to-market strategy and financial models, alongside the $25,000 in seed funding.
What Makes the Structure Different
Most accelerators work with startups that already have some traction, a working product, early users, maybe some revenue. Injini’s programme is deliberately built for the stage before that even exists, backing teams that aren’t companies yet and walking them all the way to a tested product rather than accelerating something already built. The context lead requirement is the mechanism doing the real work here: pairing every builder with someone who has genuine, lived understanding of the education problem is Injini’s answer to a common failure mode in edtech generally, technically impressive tools that don’t actually fit how real classrooms, real teachers, and real students function day to day.
The Honest Question Still Open
Whether eight months and $25,000 is enough to turn a validated idea into a venture capable of surviving past the programme is a fair question, and one Injini’s own structure doesn’t fully answer in advance. What the programme is better positioned to answer is the narrower, more immediate question of whether these tools actually work in the classrooms they’re designed for, since that’s precisely what the context-lead requirement, the built-in user testing, and Injini’s Monitoring, Evaluation and Learning support are built to measure directly, rather than leaving it to be discovered only after launch.
