Africa's newest AI fund pays before founders pitch
Askya Investment Partners is committing at least $200,000 to one company in a six-week, no-equity-required programme built to fix the fundamentals that keep African AI startups from scaling, not to hunt for a pitch to fund.
Two hundred thousand dollars. That is the minimum Askya Investment Partners is committing to at least one company before that company has raised a hand, built a deck, or asked for a dollar. The Pan-African investment firm announced this week that it is opening applications for a six-week programme for ten African AI startups, and the unusual part is not the size of the cheque. It is the order of operations: the money comes first, the pitch comes later, if at all.
Askya's founder, Babacar Seck, is explicit that this is not an accelerator in the batch-and-demo-day sense. "You noticed we did not call this an accelerator," he told TechCabal. "It is an AI growth platform. Most accelerators are focused on fundraising ... the raise is the outcome of the programme." Seck built his reputation doing the opposite of hands-off capital: as a founding member of Proparco's $300 million venture programme and later as chief executive of Digital Africa, he helped finance Moniepoint, Jumia, GoMyCode and Complete Farmer, among roughly twenty startups he has personally backed. He says those bets have returned more than $120 million in profits, including one listing on the New York Stock Exchange. Before venture capital, he advised the chairman and chief executive of AXA on the insurer's digital transformation and emerging-markets expansion. Both his track record and his current firm's founding are independently documented on Askya's own site and corroborated in a separate June profile by iAfrikan, giving the launch a verifiable foundation beyond TechCabal's own reporting.
Seck's diagnosis of Africa's AI ecosystem is narrower than the usual funding-gap story, and that is what makes the programme worth watching. He argues that African founders routinely reach product-market fit and then stall — not for lack of intelligence, but for lack of exposure to the unglamorous mechanics of scaling: governance structures, hiring discipline, pricing strategy, distribution. Cloud providers have told him that some African startups already spend more than $30 million a year on compute, even as the continent complains it needs more data centres — a mismatch, in Askya's telling, between where the money is going and where the capability lives to spend it well. Universities train AI engineers who then emigrate, Seck says, because no local company operates at the scale to hold them. Corporates want to work with African AI startups but cannot tell which ones are credible partners. Each of these is a coordination failure, not a capital shortage, and no single funding round fixes any of them.
The programme is built around two blocks of opportunity Seck sees as underexploited. The first is voice AI in African languages, aimed at markets where a majority of people do not transact comfortably in English — Seck cites the roughly 80% of South Africans who do not speak English at home as one illustration of the gap that voice-first products, deployed over WhatsApp or ordinary phone calls, are built to close. The second is what he calls AI for the real economy: Nigerian energy-optimisation startups pairing weather and grid-cost data to right-size solar, battery and generator installations, cutting energy spend by an estimated 20 to 30%; and health-insurance fraud detection using image recognition to authenticate claims in a market where only about 5% of Nigerian adults hold health insurance, held back in part by fraud rates Seck puts at half of all claims filed.
Askya's own read is that seven to fifteen years is a more honest timeline for building what Seck calls "tech Dangotes" — durable, generational companies — than the five-year horizon most venture timelines assume. That framing, not the size of the cheque, is the real bet: that Africa's AI companies are currently rationed by exposure and infrastructure more than by capital, and that fixing the former will do more for the continent's next decade of AI builders than another round of funding announcements. Ten companies will test that thesis starting with this cohort; eligibility runs to companies with a working product and first customers, up to pre-Series A, with no minimum revenue required. Applications are open now, with partnerships still being finalised alongside Deep Learning Indaba, Africa's largest AI research community.
