In a recent CIO South Africa webinar titled retail reimagined: leveraging AI for a competitive edge, Dr Prince Abudu, head of technology for AI at Cassava Technologies and Boyd Chislett, chief business officer at Liquid Intelligent Technologies outlined a vision for how African retailers can gain a competitive edge through agentic AI and the continent’s first AI factory.
With the increasing adoption of AI across all sectors, IT leaders across Africa are under mounting pressure to accelerate digital transformation in their organisations. The June 2025 launch of Cassava Technologies’s AI factory, developed in partnership with NVIDIA, is set to expedite this process even further and place Africa at the forefront of AI innovation.
Speaking at the webinar, held in partnership with Liquid Intelligent Technologies (a Cassava Technologies business), Boyd explained the importance of the company’s infrastructure-first approach to building the AI factory. With five data centres already operational across Africa and planned expansion into Morocco and Egypt, the company is laying the groundwork for AI at scale.
The AI factory is envisioned as an industrial-scale engine for building intelligence. “An AI factory is simply data in and intelligence out,” Prince explained.
But more than just infrastructure, it’s the ecosystem approach that makes the initiative transformative. “AI in isolation doesn’t deliver full value. It’s about the full ecosystem,” Boyd added.
An AI factory can best be described as the industrial revolution for artificial intelligence. “The purpose of creating an AI factory is to industrialise the entire process of creating and managing intelligence models” said Prince.
Localisation as a strategic imperative
Localisation is a key feature of Cassava Technologies’s AI factory. “There are over 2,000 languages across the African continent. You cannot look to partner with Eurocentric and US-based organisations to localise solutions. It’s essential that we build local language models that are relevant to a particular community,” said Boyd when explaining Cassava Technologies’s strategy.
In line with the company’s localisation strategy, the commercial models of the AI factory will be priced in local currencies. “We’re mitigating foreign exchange fluctuation risks and ensuring accessibility by aligning our commercial models with local economies,” Boyd added.
One of the most exciting frontiers discussed in the webinar was the rise of agentic AI. These are autonomous AI systems that can self-operate and self-deploy. “The idea is to enable AI to have end-to-end autonomy and operate as an agent without human intervention,” Prince continued. This development marks a shift from static machine learning models to self-adapting intelligent agents, offering scalable, real-time responsiveness.
The use cases for agentic AI, according to Prince, can impact every part of a retail business and drive efficiency. “Our AI factory will have enterprise-grade NVIDIA containers, which customers will be able to access through licensing, meaning they can automatically use feature catalogues and interference services. This will enable them to spin up use cases in a matter of weeks, not months,” Prince explained.
To drive retail transformation, Cassava’s platform will offer pre-built, retail-specific use cases. “Many CIOs and CTOs have held on to the notion that they need to completely transform their IT infrastructure and data before implementing AI technologies, but Cassava’s AI factory in partnership with NVIDIA speeds up the digital transformation process. The AI factory can take the data sitting in silos, automatically create an integration layer, feed it into a model, and create use cases for retailers,” Prince continued.
Retail innovation
Agentic AI is also set to improve the retail customer experience. The use cases discussed in the webinar included visual product search and inventory forecasting, both areas where African consumer behaviour offers a distinct advantage. “Our communication is often visual and auditory, not text-heavy as is the case in other cultures. Visual AI will resonate more deeply in African markets,” Prince noted.
According to him, retail adopters will be able to measure the impact of their AI investment in as little as three months. “While you will already be able to measure impact in three months, realistically I think 12 to 18 months is where the full ROI will become visible.”
Prince concluded: “Africa has been a consumer of AI that is built somewhere else with all the biases of the places it was built in. We will finally have the opportunity to build our own LLMs that incorporate African culture, experiences and languages through the AI factory. And these innovations can be embedded directly into retail systems to drive customer engagement, revenue and operational excellence.”
