South Africa’s agricultural sector is booming. It (agriculture, forestry and fishing) accounted for about 2.5–3.2% of GDP , generated roughly R495bn in total income in 2023, according to Stats SA’s Agricultural Survey 2023 .
Digital start-ups such as Khula!, an online marketplace linking smallholders to suppliers and retailers, promise to modernise South African farming. But the field is still far from level. Data remains scattered, systems disconnected and too many farmers are operating in the digital dark.
That’s where Digital Public Infrastructure (DPI) could come in. It functions as the backbone of a country’s digital economy – the interoperable systems that make IDs verifiable, payments instant, and that facilitate data flow between platforms instead of getting stuck in silos.
At the Global DPI Summit in Cape Town, Nidhi Parekh, a digital specialist at the World Bank, said that DPI was “only as good as its use cases”. And what better test case than agriculture — a sector, she said, “that needs reform given growing food insecurity, growing population, and the fact that it employs over a billion people globally”.
The problem, according to Parekh, is that current systems are “fragmented, stagnant and dominated by a few big players who can afford the high cost of digital tools”. DPI offers a way out by creating reusable, modular “building blocks” that anyone can plug into.
The Indian stack
If one wants to see DPI in action, look east. India’s AgriStack is the poster child for what happens when scattered data is turned into infrastructure.
“It’s about building a few core registries and then enabling solutions around them,” explained Mitul Thapliyal, managing partner at MicroSave Consulting, which is one of the companies involved in implementing India’s agriculture DPI systems. He pointed out that with just three registries – the farmer registry, the geofenced land registry and the crop registry – “you can address pretty much all the use cases”.
In archetype for an API services layer or services gateway. It details common building blocks and a generalised way that data is exchanged via APIs. (Graphic: World Bank)
Launched in 2021, AgriStack connects farmers’ verified IDs to digital land and crop records. The results are quicker subsidies, digital credit scores, and hyper-local weather advisories. In the state of Bihar, the Bihar Krishi platform sends millions of advisory messages to hundreds of thousands of farmers, a report by the World Bank notes.
AI in Agriculture — The Future of Farming
Read more: World Bank and Google activate partnership to reach 100 million with AI-driven public services
India is a large and diverse country, which makes it complex to roll out a large-scale DPI initiative like this. “Agriculture, I would say, is the most, or amongst the most difficult sectors,” Thapliyal said. “It’s unorganised. It’s huge, there are so many value chains. There are so many variables. And in the segment that we’re working for, smallholder farmers, these problems are compounded even more.”
Across Indian states, he said, there had been disparity in implementation and progress. “In one state like Bihar, [the] registries didn’t exist.” Physically visiting farms and farmers to put the three foundational registries in place required a significant amount of elbow grease.
Ethiopia’s digital farmland
Like plenty of African countries, agriculture also plays an important economic role in Ethiopia. There, agriculture makes up 72% of GDP and employs 64% of the workforce, said Dr Girum Teklemariam, digital director at the Agricultural Transformation Institute. Most of the country’s 18 million farmers till less than a hectare of land.
Teklemariam explained the challenges stemming from the existing digital initiatives in the country’s agriculture sector, which he said were “scattered all over the place”.
“You cannot draw context from existing data,” Teklemariam said. “There’s no way of triangulating the elements that are scattered in different databases.” He said that every institution kept its own data, following its own formats, and that “there are no clear standards that are followed by different institutions”.
Value of Digital Agricultural Roadmaps (DARs) in India and Ethiopia. (Graphic: World Bank’s Digital Agriculture Roadmap Playbook)
To fix that, Ethiopia launched the Digital Agriculture Roadmap — a plan to give every farmer a unique ID and tie it to a national data stack with open APIs. Phase one runs until 2029, focusing on credit access, input distribution and early-warning systems for drought and pests.
Planting the digital seed
Here at home, we have ambition by the truckload. The MyMzansi digital roadmap, driven by the Presidency, aims to link every citizen to a digital ID, payment system, and data exchange layer. Agriculture could easily be the next add-on.
Our fields tell a different story. A GreenCape market brief on the local agri-tech sector notes that farmers still struggle with “difficulty accessing finance for demonstrating and piloting, and a lack of technology awareness and resistance to change”. Too few people have the tools, trust, or bandwidth to join the digital party.
If South Africa were to apply DPI thinking to agriculture, three opportunities stand out.
First, digital agri-finance. With a verified farmer ID linked to land and crop data, banks could see what smallholder farmers actually produce. In Ethiopia, Teklemariam said, “proper data gives banks the confidence to provide loans”. The same could unlock billions in rural finance here.
Second, AI advisory. Parekh noted that because data often sat across various departments, ministries and private platforms, existing systems may be giving generic, low-impact advice to farmers.
DPI could change that by combining soil data, rainfall forecasts and market prices into tailored, real-time guidance.
Third, open networks for innovation. Dipika Prasad, DPI product lead at Google Cloud, described these as “systems that let farmers access all services on an open, interoperable platform, not a walled garden that belongs to any one company”. In India, such networks have already boosted farmers’ profits by more than 10%, she said.
None of this works without trust. Farmers need to know who owns their data, how it’s used, and how to switch it off. Ethiopia has accounted for an “assisted mode”, which uses local extension agents to act as digital intermediaries to help farmers with internet access and online tasks, due to the country’s low digital literacy.
The country also has a hotline that operates as an extension service, providing advice to farmers through interactive voice responses and SMS systems.
South Africa’s advantage is that its digital governance framework already exists at the national level. The challenge is to make agriculture a priority layer in that stack and to ensure that no farmer is left behind in the process.
DPI won’t solve every problem, but it can give all farmers the visibility, access and agency they’ve lacked for decades.
