In April 2020, Togo had no functioning social registry and a population suddenly cut off from daily income by pandemic restrictions. Within weeks, the government launched Novissi, a cash transfer programme delivered entirely by mobile money. Eligibility was first determined using the voter register, which covered most adults. Later, working with researchers at the University of California, Berkeley and the nonprofit GiveDirectly, Togo used satellite imagery and mobile phone data to target the poorest areas. Recipients enrolled by dialling a short code. No offices, no paper, no queues.
Novissi is usually told as a story about crisis response. It is better understood as a story about innovation through reuse. Togo did not invent new systems. It combined existing ones, a voter database, mobile money networks and phone data, in a way nobody had tried before. That is the promise of digital public infrastructure: once the foundations exist, new services can be assembled from them quickly and cheaply.
Much of Africa’s emerging foundation is open source, and that choice is shaping who gets to innovate. MOSIP, the Modular Open Source Identity Platform developed at the International Institute of Information Technology Bangalore, now underpins national ID programmes in Ethiopia, Togo, Sierra Leone and Morocco, among others. Because the code is shared, governments avoid being locked into proprietary vendors for decades, a trap several African countries fell into with earlier ID projects that cost hundreds of millions of dollars and delivered little.
The same logic applies to payments. Mojaloop, open-source software for connecting banks and mobile money providers, sits behind Rwanda’s eKash switch and informed the Bank of Tanzania’s instant payment system. The aim is simple and difficult: let money move between any two accounts, instantly and at close to zero cost. Cheaper transfers matter most to people who move small sums, often several times a day.
Open, documented interfaces change the economics for developers. When a national system publishes how others can connect to it, a small team in Kampala or Dakar can build a school fee tool, a pharmacy supply platform or a farm subsidy app without negotiating with dozens of banks and ministries. Rwanda’s Irembo portal, which now bundles more than a hundred government services, grew precisely this way, adding new services without rebuilding its core.
Global momentum is building behind the approach. The 50-in-5 campaign, launched in 2023 by a coalition including the UN Development Programme and the Gates Foundation, aims to help 50 countries design and deploy DPI within five years. Sierra Leone, Ethiopia and Togo were among its first members.
But reuse has limits that practitioners rarely advertise. Open-source software is free to download, not free to run. Deploying, securing and maintaining national systems demands engineers that many African governments struggle to hire and keep. Systems launched in emergencies often skip safeguards and never add them later. And some governments adopt open code while running the resulting platforms as closed projects, which defeats the purpose.
Ownership remains an uncomfortable question. Much of the funding, technical advice and core code still originates outside the continent. For DPI to become a genuine engine of African innovation rather than an imported template, more of the design and engineering must happen in African universities, startups and public agencies. There are encouraging signs, including regional training programmes for DPI engineers and procurement rules that require local participation.
The lesson of Novissi is not that Togo got lucky. It is that countries with even partial digital foundations can respond, and innovate, far faster than those without. The next breakthroughs in African public services will come less from invention than from how imaginatively existing systems are combined.
