[ Electronic Tone ]
[ Music ]
>> Maria Varmazis: Welcome to "AWS in Orbit." I'm Maria Varmazis. We're working with AWS to bring you an in-depth look at the transformative intersection of cloud computing, space technologies, and generative AI. On "AWS in Orbit," we're exploring not just what's possible, but what's meaningful in the realm of space and cloud innovation. We grapple with the complex challenges and unparalleled opportunities that arise when we use space to address pressing issues right here on Earth.
[ Music ]
Episode 4, Degas in Ghana, empowering small farming through AI and space technologies.
[ Music ]
Yohei Nakayama, chief technology officer at Degas Limited, shares how his company is revolutionizing small farming in Ghana through AWS's cloud, AI, and space technologies. In this compelling conversation, we learn about Degas's innovative use of generative AI to promote regenerative agricultural practices. Yohei provides insights into case studies that demonstrate the transformative impact on local farmers and the environment. First, let's hear from Yohei.
>> Yohei Nakayama: Hello, yes. I am Yohei Nakayama, CEO of Degas. And my background is a little bit research-oriented. I started my research career in the space technology field. During my PhD course, I used a supercomputer to simulate the space environment, and verify its accuracy using satellite observations. And after I got my PhD, I shifted my field a little bit, and joined AWS in 2017. And this was because I felt the profession of cloud computing at AWS, I mainly worked in the machine learning field. So as a data scientist, I developed machine-learning solutions in a wide range of domains, and it was a great opportunity for me. I learned a lot how to use ML for [inaudible] programs. But when I met Doga, founder of Degas, he introduced me to the concept of Degas, and I thought that I could utilize my background of space technology and machine learning. So I decided to join Degas as CEO. At Degas, I'm using satellite observation and machine learning techniques to help African farmers to increase their income.
[ Music ]
>> Maria Varmazis: And now that we've met Yohei, to help tell the story of how Degas is improving the lives of farmers in Ghana through AWS technology, we're going to bring in another voice now to the conversation.
>> Emma Higashikawa: Hello, I am Emma Higashikawa and I am a solutions architect on the aerospace and satellite team at AWS. And what that means is I support space industry customers in achieving their missions through designing and implementing solutions on AWS. I originally started my career in a very different field, actually as a mechanical engineer in the automotive industry. And during my time there, I witnessed what I felt were a lot of challenges and inefficiencies in the design process. And that made me start thinking about how that could be transformed by new technologies like the cloud. And so that's what prompted my transition to AWS. On the aerospace and satellite team, our team is a dedicated team that combines space industry experts with product expertise with the ultimate goal of optimally supporting the needs of our customers in the space industry. And so I'm now working in an area that I'm very excited, I've always had a passion for. And I'm extremely excited to be supporting customers like Degas today who are really tackling head-on these really significant global challenges.
>> Maria Varmazis: Fantastic. Thank you both so much for those wonderful introductions. Yohei, you did a wonderful job bringing us into sort of how you started at Degas, and what brought you there. So I thought next it would be a great time for us to talk about what Degas is doing.
>> Yohei Nakayama: Oh, yes. Let me explain Degas's mission first. So our mission is changing people's lives dramatically. So the term "people" here refers specifically to African farmers. And what we are doing is, like, Degas platform, is providing financial opportunities with farmers, and implementing data-driven precision farming. So let me explain a little bit financing part first. So it's like, in-kind financing. So specifically, we purchase fertilizers and seeds and from major agricultural companies and provide them as a loan to farmers. And after the financing part, we start the precision farming part with the farmers. So at Degas, we are hiring around 100 agents [inaudible] and the agents work together with the farmers to grow their crops in an optimized way. And because of that, the land productivity greatly increases -- eventually the income of the farmers are increasing. So that's what Degas is doing, basically.
>> Maria Varmazis: I would love to know a little bit more about specifically how Degas is achieving this mission using AI and space-based data. How does that come into play?
>> Yohei Nakayama: Yes, so basically to realize this business, we are developing two main data streams. One is coming from ground. So our agents are using our in-house Android application. And it's like boots on the ground data collection. And the main second data stream is coming from the Earth's observation. We are collecting many types of observation data, like both optical and radar satellite observation data to analyze the farmlands. And basically -- and we are -- the two data streams forms a data link on the cloud, and we are using the data link to analyze the farmland, and create machine-learning models. For example, the machine-learning models is like for detecting floods or drought or calculate farmers' credit or predicting [inaudible]. So that's how we are using technologies.
[ Music ]
>> Maria Varmazis: Let's pause for just a moment in the conversation for a bit of additional context on the technology at work here, and how it's being used throughout Africa to improve the lives of people who are being disproportionately affected by climate change. How can these somewhat abstract concepts like big data via cloud, machine learning, and space data make a real positive impact in people's day-to-day lives? Well, this is the kind of question that is exactly the domain of expertise of Zimbawbe-based space law and policy analyst, Ruvimbo Samanga. So I asked her for her perspective.
>> Ruvimbo Samanga: I think the first and foremost aspect of development that we can look at is obviously economic development, the contributions that data will have to fostering, bolstering different sectors, so for instance there's agriculture that could benefit much from the predictive analysis on crops, and how we can optimize our yields. We can also have an impact in disaster management and risk assessment looking at weather forecasting data that can help us be more prepared. There aren't a lot of weather stations in Africa that give African institutions the indigenous capacity to sort of look towards the future and prepare for natural disasters. We've seen, for instance, in 2019, there was a very devastating cyclone, Idai, which devastated about three southern African countries mainly, and over 900,000 people were displaced. And of course this might have been mitigated with more allocation of resources. Still on the economic front, we can look at data as a huge tool for decision-making not only in these different sectors, but also in organizational settings as well. The more data you have, the more you can make data-driven decisions, as cliche as it sounds. And we then see from that perspective that these solutions like Amazon Web Services that give us an opportunity to put all of this data in one place, in an actionable place, and on these platforms where it's also easily shared across different users is also very hopeful for growth of institutions and organizations as a whole.
>> Maria Varmazis: Back to Yohei now, for more on the challenges farmers are facing in Ghana, specifically.
>> Yohei Nakayama: Yes, so basically when I come to Ghana, the mobile network is not stable, and electric system is not stable. And how to access, like a high-quality [inaudible] which is a little bit lacking. So we are developing farmers' credit to provide financing opportunities so that's one challenging part. And how to scale our business, which is also challenging. So all that we are developing our in-house [inaudible] as I stated, and for that, we are collecting tons of data to calculate the credit.
>> Maria Varmazis: Excellent. Alright, so we're talking about large amounts of data. How is that being processed? Is that something that Degas is doing? Is that something that you needed help with, or what's happening there?
>> Yohei Nakayama: Mm-hmm. So for example, there is one application called Farm Mapping. So the [inaudible] for example, farmer agent go to the farmland with the farmer, and walk around the farmland. And then the Android application collect the series of location where the agent worked. And we can get the figure of the farmland. And the data is sent to the cloud. And on the cloud, we map past observation data to the polygon and calculate for example, vegetation index to check the farmland and the stages of the farmland. That's how we're merging local data to observation data.
>> Maria Varmazis: So let's talk a bit about the big picture. The motivation for Degas to be in Ghana, to be in Africa, and help solve these incredibly difficult challenges, like let's set the stage and get a sense of what are the challenges there, especially in the face of climate change?
>> Emma Higashikawa: Yes, so Africa compromises approximately 17% of the world population, although it only contributes actually less than 4% of global greenhouse gas emissions. But despite that, Africa is already contending with some of the harshest impacts of climate change. That includes extreme heat, you know, drought, desertification, flooding, and that really heavily impacts, especially agriculture and food production. And this is especially devastating in Africa, because more than 60% of the population of sub-Saharan Africa are smallholder farmers. That means a majority of the population depends on agriculture for their livelihoods. And of course, this contributes to growing food insecurity, displacement, and conflict in the region.
>> Maria Varmazis: There's a phrase that came up a number of times about regenerative agriculture. Yohei, could you walk me through what regenerative agriculture means?
>> Yohei Nakayama: So originally, every country has a lot of, like, definitions, but basically it's a way of agriculture which can issue carbon credit. So it's -- it has two meaning, which is -- it's sustainable. Sustainable way, and at the same time, it reduce CO2 emission. So we are doing regenerative agriculture with the farmers, and it's always difficult or complicated way, but using our Android application and the chatbot, we are expanding this regenerative agriculture in Africa.
>> Maria Varmazis: Wonderful. Okay. So we have this incredibly sophisticated chatbot that runs on a large amount of data. It also runs on an AWS service. Emma, could you walk me through a little bit about that, please?
>> Emma Higashikawa: Yes, of course. That's correct. So the chatbot that Degas is currently building runs on Amazon Bedrock, and Amazon Bedrock is a service AWS recently announced that lets customers easily build and scale generative AI applications [inaudible]. And the first thing is through Bedrock, AWS offers access to a number of high-performing foundation models, or FNs, and all this through a single API. Currently Bedrock provides FNs from Amazon as well as from leading AI providers like AI21 Labs, Anthropic, Cohere, Meta, and Stability AI, all through a single pane of glass API. And there's two I think key features of Bedrock that Degas has really leveraged in the solution we are discussing today. And the first one is the ability to easily experiment with and compare the performances of different foundational models for specific use case through things like playgrounds and the single pane of glass API. The second is the ability to create custom solutions integrating existing AWS services, and also using your own proprietary data in AWS with techniques such as retrieval-augmented generation, or RAG. And especially to the second point, I think Degas's solution is very exciting, because it leverages the power of other AWS services like Kendra for searching scientific papers and agricultural journals. Also retrieving, like, farmer metadata from RDS, an RDS database. And also Athena, to leverage and search Degas's existing dataset in S3.
>> Maria Varmazis: Yohei, is there anything you wanted to add to that about how Degas is using AWS?
>> Yohei Nakayama: Yes. When it comes to the better points of Bedrock, I want to add one point, like it's an API-based service. So we don't have to manage the infrastructure. So we can easily use that without any, like, management. That's really helpful for us. And when it come to the chatbot, yes -- or how to scale precision farming and regenerative agriculture which is really important, but these two methodologies are really complicated. So how to teach, or how to tell the way is really important, and for that, we are using the chatbot. So with that, we don't have to hire so many agronomists, right, in Ghana. It's really costful [phonetic]. So we can expand this business without heavy cost. That's also a really important part. And how the app is working is that for example, the agents will ask several questions related to precision farming and regenerative agriculture. For example, the question will be, like, what's the distance, best distance between planting [inaudible], or should I do [inaudible] at this time, that kind of thing. So that's the main question. But the answer is -- answer depends on the situation or the timing. So our own cloud, we are checking the question, and at the same time, searching document, our protocol, or research paper. And also checking the farmland status, which is also collected and saved on our data link. And not only referring to the question, but referring to the condition of the farmland and our protocol, we're providing answer to the agents. So that how we are operating the chatbot.
>> Maria Varmazis: I wanted to talk a little bit about the credit-scoring model. Yohei, could you talk a little bit about that model?
>> Yohei Nakayama: Oh, yes. So yes, there are a lot of data points. With that, we are calculating a farmer's credit, which is our core technology. Because with that we can, like, provide financing opportunities to the farmer. And how we calculate -- it's totally relative to their yield, right? Because it's directly related to their income. The credit model consists of for example, [inaudible] model land risk. For example, predicting flood or drought, scoring the distance of the fees or checking how to apply our fertilizers to the farmland. So all that kind of things are also digitized and collected. And we are calculating credit using the ground data and also observation.
>> Emma Higashikawa: This is somewhat indirect and perhaps lesser-known, but data from space and satellites is really integral to understanding and forecasting weather and climate. For example, there's what's called like, the GlobalL Precipitation Measurement, or GPM, which is a joint mission between JAXA, NASA, and a number of other space agencies. And what they do is, they operate a constellation of satellites that constantly measures ground precipitation, atmospheric moisture around the entire globe. And that's not something that can just be captured by a land-based sensors. And the data from GPM really informs our understanding of global rainfall distribution and the global warming cycle. And in light of Africa and climate change impacts to Africa, that space data is especially important where, you know, we're seeing increased degrees of extreme weather events, droughts, storms, devastating rains. You know, not just in Africa, but also all around the world. So space data, let's say, is impacting our lives on Earth. With regard to the credit scoring model, I did want to mention, to grow and support farmers in Africa, while also contending with climate change, farmers need raw materials, farming supplies and routes to market, and in the context of loans and support in the US, we have things like -- measures like credit scores that determine things like the loans that we can get or the mortgages that we can get. But of course, the same model can't really be translated to smallholder farmers in Africa. And so what I find exciting what Degas is doing here is developing this AI-based credit scoring model that helps determine the success and compliance of farmers in the program to, you know, determine financing, and also create a positive growth cycle.
>> Maria Varmazis: Absolutely. I think that's very valuable context, Emma. And I think you very much for that. I think that's very helpful.
[ Music ]
>> Yohei Nakayama: When it come to the market impact, we just did [inaudible] organization to create an impact report. And according to the report, 90% of contracted farmers saw some increase in their income. And 48% of them saw a significant increase in their income. Regarding to the quality of life, the report stated that the quality of life of 97% of contracted farmers improved. So even for, like, a marketplace perspective, we think we are impacting to the farmers. And [inaudible] to the one specific success story, I heard that one of our contract farmer's children was able to go to university. Yes. Because they could increase their income with datas, and they could manage to pay initial fee. Like, much of this is due to his efforts, but going to university is very expensive, right? So we could eliminate the financial barrier, which was a milestone for us. And also, I think around two years ago, we received letter of appreciation from the president of Ghana, Nana Akufo-Addo. So we are thinking that we found that our business is working fairly positively in Ghana.
>> Maria Varmazis: I would love to know the long-term vision that Degas has for -- maybe its plans or its impact or hopes for the future, even.
>> Yohei Nakayama: Yes, so one of our goal is developing or forming economic zone on our farmers' network. Before like, local microfinancing companies could not touch smallholder farmers in Africa. But we are increasing their income as well as calculating credits, their credit. So by providing the farmers credit to, for example, [inaudible] microfinancing companies, we will be able to provide not only [inaudible] guaranteed financing, but also various financing opportunities such as their children's students loans or motorcycle loans. So that's our next goal. My vision is that using space technology for peoples on Earth, it's not only African smallholder farmers, but I think one of -- even for [inaudible] language model, many scientists are using that. But how to use is also important. So I think that I want to use ML technique or space technologies for something like [inaudible]. So for now, I add data so I'm using these technologies for increasing these farmers' income. So my vision is now after increasing their income, but improving their quality of life, or like, creating economic zone, or something like that. That's our -- my vision, or why I'm doing this.
>> Maria Varmazis: this is a wonderful collaborative effort. I love that. Emma, you have a really interesting perspective. I'm curious what you think about where these technologies can go, how many customers are using them, and where you see them taking this.
>> Emma Higashikawa: Wow, that's a big question.
>> Maria Varmazis: I like big questions.
[ Laughter ]
>> Emma Higashikawa: One thing I can definitely say without a doubt is that I am constantly surprised and excited and really blown away by the innovation and the ideas that come from our customers, and that's including Yohei-san and Degas and what they're doing, what they're seeking to do with the technologies. To be honest, AWS is a provider of technologies, but really the innovation happens at the customer end, and so that's what I want to support. So support the innovation of our customers, ideas of customers that have really fundamental impacts on the world, and that's where I get my, you know, everyday excitement to go to work. And that's where I see really the synergy between AWS services and ideas that our customers bring really moving forward [inaudible].
>> Maria Varmazis: Yohei, Emma; both of you. Thank you so, so much. I really, really appreciate your time and your incredible insights. Thank you both so much.
>> Yohei Nakayama: Yes, thank you, Maria.
>> Emma Higashikawa: Thank you.
[ Music ]
>> Maria Varmazis: And that's it for "AWS in Orbit" Episode 4, Degas in Ghana, empowering small farming through AI and space technologies. A special thanks to Yohei Nakayama, Ruvimbo Samanga, and Emma Higashikawa for joining us. For additional resources from this episode, and for more episodes in the "AWS in Orbit" series, check out our show notes at space.n2k.com/aws. This episode was produced by Alice Carruth and powered by AWS. Our AWS producer is Laura Barber. Mixing by Elliott Peltzman and Tre Hester with original music and sound design by Elliot Peltzman. Our executive producer is Jen Ivan. Our PD is Brendon Karpf. And I'm Maria Varmazis. Thanks for listening.
[ Music ]
