Togo and UC Berkeley: A data science alliance to build a dynamic West African nation

Forged in the crisis of COVID, a relationship between the small but innovative country and the Center for Effective Global Action has blossomed into a vital collaboration. Leaders foresee advances in agriculture, health, education and other areas. By Edward Lempinen July 25, 2025 Farooq Sanni and Gnouyaro Sogoyou grew up in Lomé, the vibrant oceanside capital of Togo in West Africa, and though they did not know each other, their lives followed a similar path. Both left home and studied in France. Sogoyou stayed there, while Sanni settled in Montreal. Both are data scientists, a career they say they could not have pursued in Togo. But the young men sometimes dreamed of going home, and so they followed the LinkedIn page of Togo’s Ministry of Digital Economy and Digital Transformation . Last year, they saw a notice: The ministry, in partnership with the Center for Effective Global Action (CEGA) at UC Berkeley, was looking for two data scientists to become pillars of the new Togo Data Lab, an ambitious initiative to shape data-based policy for economic and human development. After an intensive interview process involving dozens of candidates from many countries, Sanni and Sogoyou got the jobs. An intensive training session followed in California in early 2025, and today they are back home in Lomé, doing work that they believe could change the future of their country. They are representatives of a deepening relationship between Togo and CEGA, a partnership based on advanced data science that aims to produce human benefits in areas such as agriculture, health and education. “At its core, the Togo Data Lab reflects our ambition to embed scientific rigor and technological innovation at the heart of governance,” said Minister Cina Lawson, who leads Togo’s digital transformation initiatives. “From using satellite imagery to identify unelectrified communities to supporting ministries with predictive models, the lab continues to expand the frontier of what’s possible for digital public policy in Africa — with CEGA as a trusted partner in that journey.” CEGA Executive Director Carson Christiano shares that commitment to data for the public good. “The more we can bolster the capacity of African institutions to generate and use evidence locally,” Christiano said, “the more likely we are to reach our ultimate goal, which is policy change — evidence-based policy change. The Togo Data Lab is our most innovative and promising effort yet to build capacity in a way that gets to the heart of our mission.” The Data Lab is not yet a year old, but under Lawson’s direction, Sogoyou and Sanni are developing a project with Togo’s Ministry of Agriculture to use satellite images, ground data and artificial intelligence to support the nation’s farmers in their work. The lab had a presence at the Food and Agriculture Organization’s Global Dialogue on AI in Agriculture in Rome this spring. Already, Lawson and Christiano say, the influence of the data science initiative is radiating out to other countries in sub-Saharan Africa. “Agriculture is a large part of Togo’s economy,” Sogoyou said. “If we take the wrong decisions, the impact can be very bad. We’re helping decision-makers to take good decisions, to have good impact. So it’s an honor for us to say that we are working to build something that can help people in our country .” ‘A long-term partnership rooted in innovation’ Togo is a small Francophone country, a slender wedge between Ghana and Benin on the Gulf of Guinea. It’s one of 44 UN-designated Least Developed Countries , and its 9.3 million people are among the poorest in the world. But its economy has been growing robustly for two decades; in 2023 alone, its gross domestic product surged by 6.4%. CEGA was founded in 2008 with a clearly stated mission: Support the use of evidence to inform anti-poverty policy in partnership with low- and middle-income countries. Over 70 African scholars have held CEGA training fellowships, and some of its alumni founded the Network of Impact Evaluation Researchers in Africa . During the early weeks of the COVID-19 pandemic in 2020, Lawson and other Togolese leaders recognized that their citizens could suffer devastating economic harm. Lawson’s ministry reached out for guidance to MIT professor Esther Duflo , who won the 2019 Nobel Prize in economics for her research in poverty alleviation. In an interview, Lawson recalled that Duflo “encouraged us to work with the team at UC Berkeley, particularly CEGA, because they’re known for their expertise in artificial intelligence and rigorous impact evaluation.” “The policymakers are extremely hard-working,” CEGA faculty co-director Joshua Blumenstock said at the time. “They’re staying up to talk to us at 3 a.m. in Togo because they want to be sure that their fellow countrymen don’t starve.” Togo and CEGA then teamed up to use artificial intelligence — machine learning, satellite images, data from mobile phone networks and other data — to pinpoint areas where needs were most urgent, and even individual households most in need. The government then delivered modest cash transfers to those citizens via digital networks. Five years later, Lawson is on CEGA’s high-powered Advisory Board . “What started as a data-driven emergency response,” she said, “has grown into a long-term partnership rooted in innovation, evidence and shared learning.” A dynamic new tool for complex policy challenges The technology that makes the Togolese initiative possible is called MOSAIKS , short for Multi-Task Observation using Satellite Imagery and Kitchen Sinks. The powerful tools were developed at UC Berkeley in the early 2020s, capable of analyzing streams of satellite data to create insights about such critical areas as soil and water conditions, housing, health and poverty. Despite its power, Christiano describes MOSAIKS as “lightweight.” It’s practical, relatively easy to operate and can run on a laptop. “This tool is meant to lower barriers for folks who don’t have the technical experience or expertise to build a deep-learning model from scratch,” said Sean Luna McAdams, a senior program manager who serves as CEGA’s top technical liaison with the Togo Data Lab. Typically, Luna McAdams explained, such models have required vast computing power and complex programming, and even when they were up and running, they would only focus on one variable — forest cover, for example. But MOSAIKS is much less expensive and it can assess multiple variables, making it far more versatile. Luna McAdams helped to lead the intensive two-week training early this year that brought Sanni and Sogoyou to California. The chief economist and a top technology officer of Togo’s Ministry of Digital Economy and Digital Transformation were there, too, further demonstrating the nation’s commitment. In a matter of months, the team at the Togo Data Lab has expanded to more than 20 data scientists. Berkeley researcher Tamma Carleton helped to create MOSAIKS. “We always hoped that the technology would take off outside academia,” she said. “The Togo Data Lab has made that hope a reality.” Added Christiano: “There’s immense potential in Togo — now that we have this infrastructure, the people and the networks in place — to bring in new tools and apply them to new policy challenges.” ‘So that no community is left behind’ The Togo Data Lab has joined with the country’s Ministry of Agriculture for its initial projects. In Luna McAdams’ view, that’s a “natural” first move: Food security is a core challenge in most low-income nations, and in Togo, some 70% of GDP flows from the farm economy. A lack of data is often a barrier in low-income countries, and Sogoyou and Sanni said that some of their early work with colleagues in the Department of Agriculture has had to focus on finding and sharing data from past studies. Now they’re moving forward, looking at land use, crop yields and soil fertility. In the future, they may be able to use geospatial data to assess how weather will affect the nation’s crop yields or, more ambitiously, predict bad harvests well before they devastate rural communities. “You may have a farmer,” Sanni explained, “let’s say in the north part of the country, who is growing cassava. But with our model, we realize that cassava is not suitable for your plots. It’s better maybe to grow maize. “Sometimes the decisions (about planting) are made blindly or with very partial information,” he said. “But with the satellite imagery that we have, with data available from the ministry — we give them as much information as we have, and with this information, they will make better decisions that will help food security in the country. ” In Lawson’s vision for the near future, agriculture is just a start. Data science is helping policymakers to bring electricity into remote areas and expand public lighting, and to assess classroom overcrowding and guide investments in school construction, which can improve learning conditions and reduce dropout rates. How does this work play out on the ground? Health is another focus area for data science. “By analyzing population distribution and existing infrastructure,” Lawson explained, “we used optimization algorithms to pinpoint priority areas where new health centers would significantly reduce access barriers. This data-driven approach ensures a more equitable deployment of health services, so that no community is left behind.” African strength, ‘now and forever into the future’ Perhaps just as important as the technological innovation is the innovation represented by the relationship. In the past, and even now, institutions from affluent countries often have regarded Africa with a sort of paternalism. It might be unintentional, but such relationships undercut African strength. CEGA and Togo have set out to do something diametrically different. CEGA’s relationship with Togo is “experimental,” Christiano said. The center’s work with the Togo Data Lab “is designed to help decision-makers themselves build better, more robust data systems and infrastructure, so they can collaborate on research, or generate evidence themselves, to steer more effective policymaking — now and forever into the future.” Lawson agreed, saying that a lasting relationship based on the old models would be impossible. “In the past,” she said, “too many partnerships with African governments have followed a top-down model, where expertise is assumed to come from outside and local voices are marginalized. That approach is not only outdated, but also ineffective — and frankly, it’s no longer acceptable. “What makes our collaboration with CEGA different is that it is truly grounded in mutual respect and co-creation. From the onset, our teams have been in the driver’s seat — defining the policy questions, shaping the use cases and developing the tools. CEGA has never positioned itself as the ‘solution provider,’ but rather as a technical partner, helping us build the internal capacity and systems to lead our own data revolution.” Shaping the future of African development Already, the influence from Togo’s data science initiative is radiating to its West African neighbors and beyond to other countries in sub-Saharan Africa. Christiano says that, in time, it could be a model for other CEGA partnerships on other continents. To demonstrate Togo’s influence, Lawson cites its open-source geoportal, featuring geospatial and administrative data that support planning in such areas as land use, infrastructure, education and health. Neighboring governments are already interested. But to understand Togo’s potential at home and abroad, she points to young scientists like Sanni and Sogoyou, and to others on a similar path. “We are seeing the emergence of a vibrant, skilled and growing community of Togolese data scientists, both in-country and in the diaspora,” she said. “They are becoming ambassadors of this vision ... and showing that African data talent can shape the future of governance and development.”