LocationsHow a visit to Kigali’s UniPod reveals the everyday work of building local capability
Made on Site at the UniPod
June 11, 2026
From the terrace, Kigali opened across the hills. Roads crossed the city and dropped into the valleys below. New buildings rose beside older neighborhoods. Between the houses, the roads and the lower ground, strips of green were still visible.
Down in the valleys, the city is changing in quieter ways.
Wetlands that once flooded during the rainy season are being reshaped through drainage, green spaces, and new infrastructure. Water channels follow the lower ground. Roads connect areas that once felt separate. From the terrace, you see Kigali’s shape. Closer up, you begin to see how the city works.
The University Innovation Pods (UniPod) revealed itself in a similar way.
The visit started with a formal walkthrough. The director explained what it takes to keep a place like this running: programmes that last beyond short grant cycles, financing that can support more than one budget period, research that remains connected to practical use, and students and faculty who stay engaged long enough for ideas to become working things.
We walked through the labs and workshops: design, fabrication, energy, agriculture, textiles. Each room had its own tools and work in progress. Some machines were quiet at that hour. Others looked recently used. The visit gave us the layout of the UniPod.
The rest came later, over coffee on the terrace. I asked Manzi, one of the programme leads, what success looked like for the UniPod. He paused, smiled, and said something close to: “Hmm. How do I frame that?”
Manzi began with the everyday evidence of the place.
Students were working on more than sixty final-year projects. The UniPod was hosting over thirty-seven active programmes. Each year, more than one thousand young people come through the space, learning, testing, building, failing, adjusting, and trying again. Three professors were actively mentoring students. Eight specialized labs were operating with student leadership through Innovate Impact, the campus association, supported by electronic access control.
The labs we had seen quiet in the morning were, he explained, often used late into the evening and sometimes overnight, depending on the projects under way. We also saw members of the Rwanda National Police using the design lab, a reminder that the UniPod was already serving users beyond the university.
What was taking shape here was also taking shape elsewhere. The Kigali UniPod was one of more than twenty across seventeen African countries, each adapted to its own setting and running its own programmes. From a single terrace, it was easy to forget that the slow accumulation we were watching, the projects, the late evenings, and the everyday work of making, was happening in parallel in other places.
He moved from one example to another, almost as if he were walking us back through the building.
There was the pellet stove designed by a student, sent to a manufacturer in China, then brought back to compare what could be produced abroad with what might be made in Rwanda. There were rice husks being tested as an alternative fuel. There was the aquaponics system, which had already been redesigned several times after rats and unsuitable fish species disrupted earlier versions. There was a traditional Rwandan instrument adapted for music education; AI translation work linked to local languages; and a bird-audio study in the rice fields of Musanze, where a request for heavy compute had been brought back to a laptop and a smaller dataset.
Then there was the café.
It had been running on the ground floor since 2024, managed by students after public-institution rules made outsourcing difficult. The campus association received initial support, then began producing snacks to help keep the model going. Sitting there, looking over Kigali, the café stopped feeling like a side note. It was another way the UniPod was solving its own problems.
Most of this had been outside the formal presentation. Some of it would be missed in a short visit. Much of it only appeared when the conversation moved from what the UniPod had to do to what it was actually doing.
After coffee, the same rooms looked different.
The laser cutter in the design lab had made some of the panels we had seen around the building. The plasma cutter in the nearby engineering workshop had shaped metal pieces used for decoration and interior fittings. Several wooden features, including parts of the reception area, had also been made on site.
In the wood workshop, a ShopBot CNC router stood near a mitre saw and a row of workbenches. On one bench, students were assembling a chair from curved wooden ribs cut by the CNC machine.
This was not a demonstration piece. It was meant to be used.
The same workshop was producing furniture for the building itself. Over time, the ambition is to show that offices, schools, and public spaces can source more of their furniture locally instead of importing standardized pieces. In that sense, a chair in the UniPod was more than a chair. It was a test of design, cost, production, and local capability.
Other rooms told the same story in different ways.
In the agricultural lab, bottles of INZORA wine stood beside food-processing equipment and a graduate certificate from a Mission Foundation cohort. In the energy lab, a hybrid car training rig used for battery and powertrain diagnostics was waiting for a replacement part after being damaged during a student session. In Musanze, the bird-audio researcher was still working from a laptop, the larger compute he had asked for was still out of reach. In the textile lab, industrial Brother sewing machines were lined up for training, with the same Kigali hills visible through the windows.
None of these details was dramatic on its own. Together, they showed how the UniPod works: students use equipment, things break, repairs are arranged, products are assessed, and learning continues through the work itself.
A formal visit can make a place like this look like a collection of labs, machines, and funding needs. Spend more time with it, and another picture appears. The UniPod is a workshop, a teaching space, a research floor, and a student-run platform at the same time. It brings students from different disciplines into contact with tools, materials, and problems. Some of what they make serves the building. Some serve the university. Some may eventually serve the country around it.
That is why the funding question matters. It is tied to the capacity that is already being built.
The same lesson appeared elsewhere during the visit. The bird-audio researcher in Musanze and the translation work upstairs had already pointed in one direction: that compute, like materials and machine time, was now another of the things this work was bumping against. The clearest version of it came in the working session with WiredIn.
WiredIn was the team behind the Ministry of Justice citizen-law application. The first issue was technical: local hosting and a limited compute infrastructure. But the practical story was sharper. During testing, response times spiked because the hosting infrastructure lacked the capacity to handle the request volume and this is designed for citizens requesting ordinary legal matters such as marriage, tenancy, and business registration. Because the Government had ruled out cloud deployment for this type of service, the system was being designed to queue requests. Proper high compute capacity remains the ideal option for this application."
That changed how we understood the wider compute question. Across the continent, there is a familiar line that GPU capacity is underused. It is true in some places. But Rwanda showed the other side of the problem: demand can appear faster than the infrastructure around it. From the UniPod, this looked less like a separate problem than a familiar one in another form: the same effort to build capability, only here the constraint was compute rather than wood or wire.
The financing conversation followed a similar pattern. At first, it sounded like a discussion about grants and short funding cycles. Underneath was a more practical issue: translation. Capital was available in instruments designed around assets, evidence, and investment logic. Operators were doing useful work, but much of that work still needed to be translated into a form that funders could recognize, assess and support.
The point stayed with us during the evening conversation. A place like the UniPod reveals itself slowly. It matters when you arrive, who stays to talk, which room is active, which machine is running, which student has a deadline. A quiet lab in the morning may only tell part of the story. A bench with a half-built chair may tell more.
Kigali asks for the same kind of attention. From above, you see the shape of the city. In the valleys, you see how it works. At the UniPod, the presentation gave us the frame. The terrace, the workshops, and one simple question showed us the work.