While Silicon Valley is chasing ever-more-powerful AI, Nairobi-based firm Amini AI is working towards a different goal: assisting emerging economies in creating the digital infrastructure required for AI to truly function. Amini AI, which has a small team of 25 and $6 million in investment, is testing its technology on real-world problems including water availability and agriculture. The startup aspires to be more than a mere IT company. Its goal is to become “the operating system for the Global South,” according to CEO Kate Kallot. At the VivaTech trade show in Paris, Kallot told AFP that emerging economies have a great chance to concentrate on more applied AI innovation rather than basic research.
From crop insurance to climate resilience
Amini AI is already assisting with initiatives in Asia, the Caribbean, and Africa; most recently, it signed a memorandum of understanding with Côte d’Ivoire. By warning Moroccan dairy producers about climate-related threats to their water supply, the company’s platform is assisting African farmers in lowering the cost of crop insurance. However, they are only the first steps. “Data is still analogue, dispersed, and unstructured in Africa, Latin America, and Southeast Asia,” Kallot stated. “Before we even discuss advanced AI, there is a lot of work to be done to digitise these economies.” Amini is also assisting in the development of data systems in Barbados, India, Nepal, and Cambodia, all of which Kallot believes urgently need to make the transition from analogue to digital.
Huge potential, little infrastructure
Despite having a wealth of youthful, tech-savvy talent, emerging nations frequently lack the data infrastructure necessary to fully utilise it. Kallot points out that while many people in countries like Kenya and the Philippines have studied computer science and can communicate in English, few actually put their newfound knowledge into practice. A structural imbalance is the cause of that. Despite housing approximately 19 percent of the world’s population, Africa only has 1% of the world’s data centre capacity, according to a 2024 analysis by Xalam Analytics. Just 2% of African data is presently processed on the continent, Kallot said AFP. “We continue to operate in a data-poor environment,” she stated. “And the majority of the dazzling new AI tools will remain out of our reach until that is resolved.”
Local first, frugal by design
Kallot said the US-China battle over chips has had little direct effect on emerging markets for now. But she warned that nations like Kenya are already becoming a battleground for foreign infrastructure investments, with giants like Huawei and Microsoft racing to lock in strategic partnerships.
What’s urgently needed, she argues, is regional collaboration on infrastructure, such as shared data centres. ‘In the past, building critical infrastructure meant roads or hospitals,’ she said. ‘Now, it means building your data infrastructure.’
Failing to do so, she warns, risks handing over control of national knowledge systems. ‘Most AI models aren’t trained on African or Asian data. If we don’t store and process our own, we risk erasing our culture and knowledge.’
Ironically, the Global South’s tech limitations could give rise to more energy-efficient and sustainable innovations, she added. ‘We’ve got brilliant developers doing incredible things with limited resources. We just need to surface it and give them a platform.’
