In this episode, Alex Diamond, the Head of Marketing for Descartes Labs, walks us through the evolution of the GIS industry from mapping to monitoring to advanced analytics. Historically, companies have been geospatially oriented but not geospatially informed. New technologies like cloud computing, sensors, and machine learning are transforming the industry, allowing companies like Descartes Labs to produce insights at planetary scale.
Alex explains how Descartes Labs expands the way clients examine their operations. The company’s platform can not only help clients optimize supply chains and sourcing decisions, but it also informs sustainability initiatives such as land use, emissions, and deforestation. Tune in to hear specific examples of how this technology is applied across a variety of industries and future plans for the Descartes Labs platform.
The Understory Podcast is also available on SoundCloud, Apple Podcasts, Amazon Music, and Google Podcasts.
Episode Transcript:
JJ (the host of The Understory Podcast): Welcome to another episode of The Understory Podcast. Understory is a global community and a platform for innovators and innovative companies that are trying to make a huge impact in our world and trying to make our world more sustainable.
Today, we're thrilled to have Alex Diamond, who's the Head of Marketing at Descartes Labs. Alex welcome to The Understory podcast.
Alex Diamond: Thanks JJ, it's great to be here.
JJ: To start off, tell us more about your background and then we can shift to talk about Descartes Labs and its work.
Alex: My background is securely placed in the geospatial industry. I went to the University of Illinois. I got a bachelor’s degree in geography focusing on GIS, also minored in geology. Geology was my preferred choice going into college, but I ran into geophysics and some of the math involved. I heard from a friend about this new application called GIS and as soon as I started taking classes I loved it. By the end of my senior year, I had an internship at the Illinois Natural History Survey. We were doing classification of land cover types in the state of Illinois. I look back on that now, and it's a lot of the same principles that that have progressed into machine learning, supervised and unsupervised classification.
Out of school, I got a job at an earth science software company in Golden, Colorado. I worked there a couple of years and familiarized myself with a number of different earth science software applications. We sold a lot of Esri products - I think we were the Rocky Mountain regional business partner of the year. I then moved on to a company called DigitalGlobe. They are a commercial remote sensing satellite company. I was there for about 5 1/2 years as a senior product manager on the commercial side, developed a lot of systems for commercial customers, some web services for commodity companies and overall, just had a great experience there.
When I left DigitalGlobe, I started my own company called Remote Sensing Metrics or RS Metrics with my brother and another co-founder. We had an interesting idea to take the advent of new satellites that were being launched, including World View, one from DigitalGlobe and trying to take advantage of this transition from mapping to monitoring. Considering there was more capacity in satellites in the sky and you could revisit any location multiple times, we were using it to monitor activity at retailer, parking lots, looking at changes in car traffic and trying to correlate that with transactions inside the store.
I spent about 10 years at RS Metrics and then most recently moved on to Descartes Labs and took on a product marketing role. What I really loved about Descartes Labs is the to take advantage of their platform to do anything related to geospatial analytics. It provides really low-level access to do all the things that are possible in the geospatial realm. And that's where I am today.
JJ: Thank you so much for that introduction. I think you bring so much experience in this area that it's really amazing. A lot of people are learning about what's now being called ‘Climate Tech’ and really understand our planet. But Alex, you started awhile back and have seen the evolution. Let's talk about that.
You alluded to this - that some of the principles remain the same. When you started looking at GIS and all the different entrepreneurial work that you have done, building companies to now. What has been the evolution in geospatial intelligence or innovation as it relates to the various use cases to understand the Earth.
Alex: I think the concept of going from mapping to monitoring and the sensor revolution has been the linchpin of what's enabled us to monitor our world in this way. When I first started at DigitalGlobe, there was one satellite Quickbird that they operated. At the time, it could revisit every location on Earth maybe once every week or so. It really took a couple of the newer satellites with higher capacity to more frequently revisit and look at the Earth and understand what's happening in a more continuous stream of information and building time series analysis specifically.
I would say that is probably the biggest change and over the last 10 years. It's become more pronounced - there’s something like 6000 satellites up in the sky today. Most of them are not Earth observation satellites, but the ones that are, produce petabytes of data. I think it's something like 80 to 100 terabytes a day. You really need platforms that can automatically sift through that data and pull out insights. It's sort of this combination of the sensor revolution, cloud computing and then some of the machine learning technologies - deep learning, computer vision -- have all come together to give us the tools that we need to take advantage of all this data.
JJ: Thank you for that perspective. It's mind-blowing how much data are being captured and almost like the democratization of satellites and how that's kind of driving things forward.
Let's talk specifically about Descartes Labs. It’s a well-known company this is venture-backed. They raised a lot of capital from renowned investors in this space, both strategic and institutional investors. How would you characterize the focus or the priorities of Descartes Labs? How do you take some of the technology you mentioned to drive impact with these customers?
Alex: I would say Descartes Labs were geospatial intelligence companies, so we go beyond geospatial analytics. We really want to provide intelligence to commerce at our core. We were founded out of Los Alamos National Laboratory by a group of scientists, and we perform scientific analysis of geospatial remote sensing and other diverse complementary datasets - even datasets that might come from customers themselves. We use all that data to enable sustainable sourcing, best practices, including commodity price forecasting, even efficient mineral exploration. Those are just some examples.
Some of the verticals that we focus on are consumer goods, agriculture (specifically the agri food supply chain), and then also the mining exploration industry. We’re really focused on providing companies more efficiency in their operations. And then the flipside of that is the sustainability aspects that can be achieved by that increased efficiency.
The core of what we do is built with our SaaS platform. We have a geospatial processing engine that automates the analysis of data for our users, and this enables planetary scale analysis, again, with that sort of combination of cloud computing, sensor revolution and machine learning. We can do this planetary scale analysis on whole continents looking at things like deforestation or regenerative agriculture. By leveraging artificial intelligence and machine learning, we can bring insights to companies that weren't able to achieve them before. I think when you look at the advent of business analytics over the past 20 or 30 years, it really started with data sort of inside your four walls - so looking at the output of ERP systems and optimizing your own supply chain or logistics, or operations inside of a warehouse.
I think what we want to do is to use these petabytes of external data to model what's happening in the world around you and get a better picture of how the earth and the rest of the supply chain - both your suppliers upstream and downstream - impact the context in which you operate and allow you to optimize it.
A lot of these geospatially oriented companies think of the commodity supply chains. They're very reliant on location and the earth and trade across the whole globe. They're geospatially oriented, but a lot of them still aren't geospatially informed, so that's what we're trying to do. We're trying to use our platform to inform their operations making them more sustainable.
We depend on these companies to feed billions of people and to supply the materials that we need for all our modern devices. I think one of the things that the companies really want to do going forward is operate in a more sustainable way and that will help them fulfill their fiduciary duty to their shareholders by operating sustainably and having a long-term view on what's best for them and for the world around them.
JJ: Yeah, and with respect to the commodity forecasting, everybody is reading a lot about the supply chain pressure and the pandemic is making it worse. In general, there is lots of pressure both on the supply chain side and also the other kinds of responsibilities as companies become more aware of ESG and sustainability.
With Descartes Labs, when you work with your customers just on the commodity, what are some of the example questions that you help them answer?
Alex: Let’s take a step back a little bit. I think that commodity intensive industries are facing dual challenges. First, you have increasing margin compression and competition coming from all sides. We’ve all heard about the increased demand and the bounce back in demand after COVID. Some of the inflation that's causing the agrifood supply chain they have that they have to deal with.
There's resource scarcity as we progress on and on throughout the years here, resources are becoming harder to come by. Extracting them from the environment requires a more careful and balanced approach, making sure that we don't continue to add externalities out into the environment.
Finally, when we look at price forecasting and securing of commodities, there's a lot of competition now with AI and algorithmic trading. Somebody says who needs to secure thousands of tons of soybean meal or palm oil. Now they're competing with hedge funds and other commodity market participants that that are running AI models or algorithmic trading strategies, so that margin compression and competition is one of the big challenges.
The second is, and this is very welcome, there's an increasing long-term focus on sustainability and biodiversity. As I said before, externalities from some of the intensive operations we've had over the past 50 or 100 years, they're finally coming home to roost. You can see this all around us in climate change and an increase in forest fires and temperature increases and changes in weather and climate.
It’s becoming a fiduciary duty to pay attention to ESG, and factors like that they have to serve all their stakeholders, their employees, the environment, their customers, and it's a very diverse world. If they want to maintain long term shareholder performance, they need to double down on ESG initiatives, and these diverse sets of stakeholders can no longer be ignored.
What we offer are sort of a combination of two sides of the same coin in terms of commodity market forecasting capabilities and sustainability initiatives. In a lot of cases, we'll produce a map of a particular commodity on the Earth. It sounds pretty simple, but in reality, it hasn't been done before in large scale.
If you want to know where all the coffee is grown in Brazil, or all the palm oil in Southeast Asia, we have the capability to use machine learning and satellite technology to map that at a very fine-grained level of detail. Once we map all of that, we can use it to target areas of commodity forests, but then also natural forest. We can use that target to basically provide a baseline for where we should be looking for things like land use, emissions, or deforestation or reforestation. That same data set that we've used again to map the natural forest versus the commodity forest, we can then use to intersect that with weather forecasting or calculating acreage and yield, and so those same datasets can be used to predict yield and forecast prices. At the same, time they can also be used for sustainability initiatives like I mentioned before, so it's this dual capability of being able to provide both commodity market forecasting but keep it within the lens of sustainability as well, and in many cases, we can use commodity market forecasting to fund sustainability initiatives.
JJ: I really appreciate that clarification and explanation and really taking something very complex and breaking it down for our audience. You make that relationship between technology impact and how they actually do that. Talk to us a bit more about some of the other offerings of Descartes Labs. Using the same kind of technology platform, a SaaS platform, how do your customers use Descartes Labs similarly or differently?
Alex: I'll give some examples within mining exploration. At this stage we all rely on the materials and elements that mining companies produce. And like any other company, they are trying to operate in the most sustainable way possible and it's just become more and more important to them. This ESG and other factors have become focus areas for institutional investors that invest in their firms and their companies.
One of the things that we're helping them do is it becomes harder and harder to find mineral deposits or to do exploration. All the easy areas for minerals like iron or nickel anything that that a company might be exploring for have already been found. Now it's just getting harder and harder to find these deposits in a sustainable way. You can on one hand just get boots on the ground and explore everywhere. That's a little bit intensive, can be difficult for the environment.
What we offer is we’ve basically taken a couple of remote sensing datasets, global data, one is called Aster and it's a global multi spectral data set, meaning it has multiple bands of imagery that extend outside the human visual spectrum. We’ve taken that data set and basically pulled out the clearest pixel possible across the whole world. When I'm talking about planetary scale capabilities, this is one we’ve produced this global map of what we call a bare earth deposit of the clearest pixel available across the whole world. We’ve assembled that using our cloud computing capabilities - this massive data set.
We've also done one for Sentinel 2, which is a somewhat similar optical Earth observation data set. We've basically created a global composite map and so now mineral exploration companies can take that data along with software tools that we built. What used to take them hours, months, days to build the clean the data project, the data color, balance the data for just an individual project area. We've taken all of that away, so now they can just basically pan and zoom anywhere they want on the earth. And that data has already been produced for them. Then we combine that with interpretation tools and other ways for geologists to be able to understand and interpret what's happening on the ground.
It just saves them so much time for early-stage exploration activities and so we've seen a lot of success with that. I think it's great because it makes them operate more efficiently and it helps them find the best candidates that would have taken them much longer on a project-by-project basis. Now we can open up the whole world to them by producing these global datasets.
JJ: When you go to your customers, whether it's the large commodity companies or the mining companies, and I know you guys serve across a wide spectrum of clients. A lot of companies think that if they have machine learning engineers, they can do geospatial stuff and that they're able to do both compute and the analysis and everything between building the pipeline at the beginning. But it's a big endeavor.
Who do you talk to when you go to these clients? Does Descartes Labs usually talk to machine learning engineers? Is it more of a strategic engagement? Getting the business side to understand the value proposition and just evolving trends?
Alex: It can be any one of the sorts of personas that you mentioned. Oftentimes, it might be a Director of Innovation or Manager of Advanced Analytics or a Sustainable Sourcing Manager and Exploration Geologist. The beauty of our platform is that it allows you to get this low-level access to geospatial data, compute machine learning technologies and you can build anything you want from that, whereas a lot of other platforms are sort of prescribed and a little bit limited in terms of the capabilities. We're talking to managers who are on the ground that are trying to reduce costs for procurement or hedge against purchasing costs. It can be a VP of Supply Chain. It's really pretty variable, I would say.
JJ: For individuals and even students and others who want to get into geospatial intelligence. Alex, what do you recommend them to study in school or how do they find work at companies like Descartes Labs?
Alex: I think that a good sort of education within the physical sciences these days often includes the necessary computer science work that you need to have today, to be successful, whether it's within commercial or academia or even within government. Our physics majors, you're going to be exposed to machine learning techniques. You're also going to be exposed to traditional statistical techniques. Or you can get a degree in physical geography, or you can have a degree in meteorology. All of these fields, any sort of physical science or computer science, is key to make headway within this field in this industry.
We see we see a lot of diverse backgrounds. There’s a lot of communication that's required. Even traditional business degrees are valuable because you can apply all the science in the world. But if you can't communicate and get a groundswell of people following in the same direction, it's hard to make a. lot of progress. Anything related to physical science, sustainability, data science. Those are very interesting areas, and I would definitely recommend students to continue along those paths.
JJ: That's great advice.
Last question, and it's a two-part question. Now what? What are some of the continued key priorities for Descartes Labs as you think about 2022? The easier question I suppose is where can prospective clients or partners learn more about the company?
Alex: I think in the remaining part of 2021 and in 2022 we're continuing to focus on our both sustainability and operations theme – enhancing companies’ capabilities to take advantage of contextual data around them and model their environment in a better way so they can become more geospatially informed.
We’re going to look more toward opening up our platforms and expanding our partnership strategy so we can have companies that really have application specific experience take our platform and apply their unique knowledge and leverage it basically to produce solutions for their customers. Those might be software companies that want to OEM or even integrate our capabilities within their broader solution. It could be channel partners who want to produce one-off solutions for customers. Definitely a focus for us in 2022 is expanding our partnership strategy and just allowing others to build on top of this great platform that we've created.
To your second question about getting in touch - it's pretty easy these days. There's a lot of information on our website. We have a contact form on our website, but we encourage lots of people to interact with us on Twitter and LinkedIn as well. The company pages on Twitter and LinkedIn, we share a lot of news and announcements, and we love to hear from people through those platforms. There are multiple ways than getting in touch with us. Just drop us a line and we can reach out and connect and a tackle some of these challenges together.
JJ: Alex, thank you so much for coming on today and sharing more about your background, your work, the focus, and the impact that Descartes Labs is making. We welcome you to come back again and continue to raise awareness about the role of geospatial intelligence in climate and sustainability. Thank you so much again for joining us at the Understory podcast.
Alex: Thanks JJ, it's been great to join you and love the work you're doing at Understory. I really appreciate it.
Join the conversation.