A Nigerian AI infrastructure startup has launched a platform designed to function as a security layer between AI applications and the models they depend on, addressing what its founders describe as a critical gap as African and global companies rush to ship AI-powered products without adequate guardrails.
Cencori, founded in June 2025 by Bola Banjo, Daniel Oreofe and Ladipo Samuel, offers a unified gateway that secures, routes and monitors AI model requests for production applications. Co-founder Oreofe describes the company as “the Cloudflare for AI production” — sitting between AI applications and the models they call, handling security, reliability and traffic routing.
The startup emerged from frustrations the founders observed while working in operations at Nigerian startups Mytherapist.ng and Eden Life. “We saw developers plugging LLM APIs into production code and simply hoping for the best,” Oreofe said in an interview with Connecting Africa. “The specific ‘aha’ moment was the realization that personal data, user emails and phone numbers were being shoved into prompt requests with zero visibility.”
The company’s launch coincided with the rise of “vibe coding” — a software development approach where AI tools such as Claude, ChatGPT, Cursor, Lovable, Bolt and Replit generate applications from conversational prompts. Oreofe said the trend has allowed developers to ship applications 10 times faster but often with little to no security in place. As AI-generated code has flooded production environments, the industry has begun referring to the resulting gap as the “infra vacuum.”
“The rise of vibe coding influenced Cencori to focus on guardrails rather than another user interface,” Oreofe said. “We realized that if the code was being generated by AI, the security had to be handled at the infrastructure layer where the requests actually exit the building.”
A core feature of the platform is automatic failover between major AI providers. “If OpenAI goes offline, requests are seamlessly rerouted to Anthropic or Gemini, with no action required from developers,” Oreofe said.
The company has remained bootstrapped and founder-led, competing with well-funded U.S. counterparts while growing through community engagement. Banjo and Oreofe built early traction through educational content on X and LinkedIn, attracting more than 180 organic users in the company’s early months and growing past 200 without paid marketing.
“Bootstrapping with personal savings was a deliberate choice to preserve optionality,” Oreofe said, allowing the founders to remain true to a secure-by-default vision without being pushed into a growth-at-all-costs model.
Early funding conversations are now focused on enterprise-grade controls, including role-based access and advanced audit logs, with ongoing discussions involving a tier-one Nigerian bank around global compliance requirements.
The launch reflects a broader maturation in Africa’s AI ecosystem. The continent has moved beyond experimentation, with startups now building real-world infrastructure and governance systems spanning African large language models, AI-driven fintech, healthcare and public-sector pilots — a shift that increasingly positions Africa not as a passive consumer of AI built elsewhere, but as a co-author shaping how AI systems are built and governed in emerging economy contexts.
