Generative AI (GenAI) has captured global attention over the past two years, including in Kenya, due to its ability to draft content, summarise reports, and offer conversational assistance. These tools provide value for teams seeking to boost productivity and reduce administrative workloads. However, GenAI represents only one part of the broader AI ecosystem.
For most Kenyan organisations, the real opportunity lies in understanding how generative and agentic technologies complement, rather than replace, each other, and how each can be applied at different stages of digital maturity.
The effectiveness of any AI system, whether generative or agentic, depends heavily on the quality of data and workflows, according to Veerakumar Natarajan, Country Head of Zoho Kenya, a cloud-based suite of business applications for sales, finance, HR, and more.
Many Kenyan organisations face challenges such as manual processes, inconsistent data entry, fragmented systems, and limited integration between platforms. “These realities make it difficult to leap directly into advanced AI use cases. Without clean, organised, and accessible data, even the most sophisticated AI systems can produce inconsistent or misguided outputs,” Natarajan explains.
He advises that the most practical starting point is not immediate adoption of advanced GenAI models, but digitisation and automation of core processes. Tasks such as routing customer-service tickets, reconciling mobile-money transactions, managing field reports, or processing sensor data may seem modest but provide immediate, tangible value.
“They reduce errors, improve consistency, and create a clearer picture of information flows. As these processes stabilise, they naturally highlight areas where AI can make a real difference,” he says.
Once these foundations are in place, AI becomes especially powerful. GenAI helps teams create and be productive, while agentic AI helps organisations act efficiently. It proposes actions, verifies them, and executes based on predefined business rules.
This distinction is critical in sectors like BFSI and public services, where trust, compliance, and accountability are central. For instance, an agentic AI-powered loan approval system may recommend an action but only execute it after confirming KYC rules, thresholds, and documentation verification. This combination ensures fast, reliable decision-making.
As Kenyan enterprises adopt AI-enabled systems, another layer emerges: context. Global AI models often struggle with local regulations, business practices, cultural norms, and sector-specific terminology. “Contextual AI and sovereign LLMs—fine-tuned with local data—ensure insights and actions reflect Kenyan realities,” Natarajan notes.
These models complement global systems, adding local intelligence required for accuracy, relevance, and regulatory alignment. Beyond technology, AI presents an opportunity to strengthen “transnational localism,” where global technology fuels local innovation and economic empowerment.
No-code and low-code tools embedded with AI allow SMEs, NGOs, and governments in regions like Kisumu, Eldoret, and Turkana to build automations without specialised expertise. A micro-insurer can automate risk assessments, a county office can streamline citizen services, and an agritech start-up can manage farmer support workflows. This decentralises digital innovation to communities that understand their challenges best.
Natarajan advises leaders charting their AI journey to start with workflow automation, build strong data foundations, introduce GenAI for productivity improvements, and gradually adopt agentic AI for secure, auditable automation. Contextual and sovereign AI models then add essential local relevance.
Kenya’s AI future, he cautions, will not be defined by a race toward the most advanced model. Instead, organisations that invest in quality data, well-designed workflows, and responsible AI systems will achieve better customer experiences, reduced costs, and empowered teams.
Embrace AI not as a flashy tool, but as a dependable partner delivering lasting impact,” he urges.
By Michael Omondi
