Generative AI for Personalized Agronomic Advice

Title

Generative AI for Personalized Agronomic Advice

Content

Farming and AI in Africa Apollo Agriculture 0 How AI is Transforming Smallholder Farming in Africa Smallholder farmers produce most of the food consumed across Africa, yet they face major hurdles—from accessing financing and quality inputs to navigating unpredictable weather and markets. In this BBC interview, AI expert Benjamin Njenga shares how artificial intelligence is helping farmers in Kenya and Zambia overcome these challenges, with tools for credit scoring, crop monitoring, and training. Farmers themselves weigh in, revealing how AI is already transforming their productivity and livelihoods. Charles BBC: Small-scale farming is a big deal in Africa. In fact, around 70% of the food we consume is produced by these farmers. They face many challenges, such as where to find the right inputs, identifying fertile land, what and when to plant, and even where to sell their produce. It's a lot to deal with. But artificial intelligence, or AI is promising to solve some, if not all, of these challenges. But with little technology being available to most smallholder farmers on the continent. Is this a viable solution? I called up Benjamin Njenga from Kenya. He is an AI expert helping create solutions for farmers in Kenya and Zambia. Sometimes, when we mention AI or artificial intelligence, it can get quite confusing. You know, people can't quite put a finger on what's being talked about. What is AI all about? Benjamin Njenga: In a simple statement, AI is the use of computer machines, enabling them to perform tasks that will typically require human thinking. Just a quick example. In ag space, which is the industry where I am in, you can be able to use a computer to read the weather data that advices farmers when is the best time to plant, which ideally this is experience that usually farmers will be able to get through experience for many years on when they usually plant. Charles BBC: So tell us about then how you are applying this technology to farming in Africa. You've talked about being able to check weather systems. What have you built and how does it work? Benjamin Njenga: As we know, smallholder farmers are hard to reach. They don't have any financial records. They are very rural and very fragmented. So one, we have been able to build credit models in terms of providing instant credit decisions to African smallholder farmers, who typically won't have any financial records for them to access financing from any traditional banking institutions. Charles BBC: How have you been able to do that exactly? Benjamin Njenga: So, you know, Charles, today in most African countries, and even if you look at smallholder farmers, smartphone penetration is still going up. But we still have many of our smallholder farmers still with feature phones. And how we have been able to do it is that we have been able to build a field officer network that comes in between, because smallholder farmers still want to interact with the physical touchpoints; they want to talk to a person, and they want to build that trust. Also, some of these farmers, majority are still not tech literate. For example, the credit model. So this farmer has no credit history as I mentioned. So using AI we are able to collect alternative data sets- data from the remote sensing, data from the farmers, data from third-party systems like the CRB and be able to use AI to be able to create credit profiles for these customers so that we can be able to make a lending decision. Charles BBC: What are some of the things that your AI tool is able to do for their farmers? If you could mention 2 or 3 things. Benjamin Njenga: So one is credit that I mentioned. The second one is training. We are able to do financial literacy training. We are able to do better product recommendations for the farmers. So we have been able to build AI systems for our field officers- and this is able to provide the extension service that the farmers need. And finally, we also use AI is so that we can also be able to train crop models that are able to monitor the growth of the crop over the season, and that is able to enable us to be able to adjust the loan deadlines at the end of the season in case there is a delay times of harvest, and that they enable the farmers to be able to get a deadline that matches when they have been able to harvest their crop and be able to sell their produce, and be able to get the return that they can be able to pay back their loans. Charles BBC: So I'm wondering whether you are having to overcome some barriers of perhaps cultures and what we are used to, because you are going to a farmer who is basically not much technology that's being employed to be able to farm, but now you are telling them they should trust maybe some information that comes through a mobile phone. How do you convince them that this will change the way they do farming? Benjamin Njenga: The most important thing with farmers changing behavior is trust. You have to build a relationship. You have to build trust. And mostly, this trust and the relationship has to be built with physical touchpoints. So, farmers are still worried about changing any behavior using text or either over the phone. And that is why we have invested in this field officer network. Charles BBC: Are there barriers to this and what does it take really, for a farmer to be able to use AI generally? Benjamin Njenga: For smallholder farmers, it’s the challenge around literacy levels, it’s the challenge around internet connection, it’s the challenge around ability to use smartphones, Charles BBC: Do you see this as a space that will continue to grow in terms of usage of AI in farming, and what would that mean on a bigger scale? Benjamin Njenga: Yeah, for sure. This is a space where I can tell you, Charles, it will really continue growing and it has massive potential. We can use AI to consume training data that could be able to guide farmers on which are the best varieties, so that they can be able to adapt to better climate-smart technologies. We find farmers are still stuck with the old technologies, climate change has come, they are still planting the old variety, but we can use AI to consume soil data, to consume training data that can be able to advise these farmers, which is the best variety for them to be able to plant in their area. Charles BBC: That's Benjamin Njenga in Kenya. We are some farmers who are using AI solutions to share the experiences with us. Take a listen. I am Caroline Kosgei . I plant maize and beans. Before starting using that technology. I used to call for seeds. Sometimes you could get the wrong seeds and fertilizers, but when we started now having a network with the technology, that is when we started having better yields, and also we could harvest, we could plant at the right time, at the right season and at least we now double or even triple our yields. My name is Flavia , and I produce maize. I'm telling you at this time with my family is getting breakfast, lunch and supper and those who are at school, I managed to take them to school. This year, I'm going to harvest maybe 50 bags of maize. Technology is good because when you are trained and given the knowledge you can know how to live.