When Ejiro Omoniyi convened a small group of investors and technologists for a live demonstration of his artificial intelligence platform, the atmosphere was analytical rather than celebratory. The platform, Lykeli, is already operational and built on an offline-first architecture designed for environments where internet connectivity is unstable or expensive. Yet the discussion quickly moved beyond functionality to a broader question about Africa’s digital direction.

Omoniyi’s presentation centred on a premise that challenges dominant AI design assumptions. Rather than relying entirely on uninterrupted cloud access, Lykeli is engineered to continue functioning when connectivity fluctuates. Through its interactive chat interface, users can ask questions, explore ideas and receive detailed responses across all topics without experiencing service interruption during network gaps.

In many parts of Nigeria and across the continent, such resilience is not theoretical. Data costs remain high, and broadband penetration is uneven. For students, entrepreneurs and remote workers, brief outages can halt productivity. Lykeli’s architecture attempts to reduce that friction by designing for constraint rather than abundance.

As an engineering approach, the model is notable. Most global AI systems are optimised for scale and continuous cloud computation. Building for inconsistent bandwidth requires a different set of trade-offs. That shift in design philosophy positions Lykeli as more than another application. It represents a deliberate rethink of how AI might function within Africa’s infrastructural realities.

A continent of constraints

Omoniyi described what he sees as a blind spot in mainstream AI development. “If systems are built only for always-online environments, they exclude people by default,” he said during the session.

The argument resonates in regions where connectivity cannot be assumed. Observers acknowledged the technical merit of a system capable of shifting between connectivity states without shutting down entirely. In a market dominated by cloud-first giants, building for low-bandwidth conditions signals independence of thought.

Yet recognition of innovation does not remove the harder questions.

The bigger question for African AI

Africa’s technology sector is vibrant. From fintech to agritech, startups across the continent are deploying AI to address local challenges. Critics, however, caution that launching new platforms does not automatically resolve structural barriers such as power supply, broadband access and device affordability.

In that context, Lykeli enters an ecosystem already crowded with ambition. The issue raised during the session was not whether the tool works. The deeper question was whether Africa needs another AI platform or a more coordinated push toward infrastructure investment.

Some participants argued that offline-first systems may complement infrastructure expansion. Others suggested that sustainable digital transformation depends less on product-level solutions and more on long-term policy alignment.

Omoniyi positioned his work within that tension. He did not frame Lykeli as a substitute for infrastructure development, but as a response to present conditions. “This is about asking what can be built responsibly within the realities that already exist,” he said.

Between promise and pressure

The African AI landscape is increasingly competitive. Users gravitate toward platforms that deliver speed, sophistication and integration with global ecosystems. Offline resilience may matter deeply in some contexts and less in others. The platform’s long-term relevance will depend on its ability to balance durability with performance.

The session concluded without a consensus. Some described the approach as pragmatic and necessary. Others questioned whether architectural adjustments alone can substitute for comprehensive infrastructure reform.

What remains clear is that Lykeli represents a serious contribution to the AI conversation in Africa. Whether offline-first systems become a defining feature of the continent’s technological trajectory will depend on adoption, execution and the pace of broader digital development.