Looking back at 2025, I see a year of decisions
Not perfect conditions. Not complete certainty. Decisions to move, to build, and to take responsibility where others were still talking. USIO exists because, at some point, you stop debating whether something should be done and start doing it.
From day one, USIO was driven by a simple belief: if AI is going to matter for Africa, it has to be built in Africa. Not consumed as a remote service, but owned, governed and operated locally. That belief shaped every choice we made throughout 2025.
Building an AI factory is not a theoretical exercise. It is operational, financial, political and human. It means aligning governments, investors, partners and talent around something that does not yet exist. It means taking calculated risks and accepting that not every variable can be controlled upfront.
That reality became very tangible during the joint press conference with Synectics in Kenya, where we announced that Uganda would host Africa’s first AI factory. Standing there together made it clear this was no longer about plans or positioning. It was about commitment. Once you publicly put a stake in the ground, execution is no longer optional.
Much of the real work in 2025 occurred out of the spotlight
We worked intensively with the core team at Synectics, particularly Ola, Adunni, Gbenga, and Randy. Extended sessions, challenging questions, constant iteration. This was not vendor coordination, but true co-creation. Decisions about governance, operating models, timelines, and trade-offs were made together, often under pressure and always with a shared sense of responsibility.
Executive alignment with NVIDIA, turning architecture choices into real commitments.
At NVIDIA’s Executive Briefing Center, sharpening our view on what it takes to build and operate at scale.
Seeing the technology up close, because development isn’t done from slides.
In parallel, we worked closely with our modular data center partner, Automation. Together with Jo, Jurgen and Guy, we translated ambition into physical reality. What does it take to deploy AI infrastructure quickly, securely and sustainably? What can be standardized, and what must remain flexible? These were not abstract discussions, but concrete design choices that directly shape delivery.
2025 was also a year of movement
Workshops in London. Executive meetings at NVIDIA headquarters. Sessions at the NVIDIA Executive Briefing Center. A visit to CERN to see modular data center concepts operating at scale in the real world. Workshops with prospective partners in the Netherlands and Belgium. Too many moments to capture fully, but all reinforcing the same point: this was never theoretical work.
On-site at CERN, to visit the modular data centers of Automation.
Inside the modular data center, where operational details decide whether a factory can truly run.
Modular realities: security, access, facilities, the practical edge of delivery.
Energy infrastructure and geography, the context every long-term plan must respect.
USIO
Throughout 2025, we worked closely with the parties required to make USIO real. With financial institutions willing to take measured risk to enable AI compute at a regional level. With NVIDIA as a core technology partner, contributing architectural depth and long-term perspective to help shape the AI factory and its operating model.
In that context, our collaboration with Auke at NVIDIA deserves special mention. Auke was our rock and compass within NVIDIA throughout 2025. Not just a connector, but a sparring partner who understood both the ambition and the constraints of building sovereign AI infrastructure in Africa. His guidance helped us navigate NVIDIA’s ecosystem with clarity, realism and momentum, exactly when it mattered most.
Alongside that, we worked with specialized AI service providers to support tenants from early ideas through to production-grade algorithms. And with entrepreneurs who pushed the concept forward, challenged assumptions and helped turn abstraction into execution. What I was reminded of again in 2025 is that sovereignty is not something you declare. It is something you build. In contracts. In governance. In daily operational decisions about access, prioritization and accountability. USIO is designed around that principle.
It is not meant to compete with global cloud platforms. It exists to do what they cannot: anchor AI capability in the region itself, integrate it into local ecosystems, and ensure that value creation stays close to the people using it.
By the end of 2025, USIO had moved beyond ambition. Partnerships were aligned. Structures were in place. And the focus had shifted decisively from planning to execution.
Looking ahead to 2026
I am not interested in bigger promises, I am interested in operational reality. In a factory that runs. In workloads that matter. In talent that grows around it. And in proving that sovereign AI infrastructure in Africa is not only possible, but viable and scalable.
2026 is about delivery.
And once the first factory is running, we move to the next.
That is how you build something that lasts.
| Raymond Drielinger Founder & Owner | MDCS.AI |
