Interview with Dr. Benjamin Laken of cervest on Data Science and its Role in New Sustainable World

Hello, everyone. Welcome to today's episode of Conscious Design. I'm your host, Ian Peterman, and I have with me today Dr. Benjamin Lakin, Head of Innovation at Cervezt, a company focused on environmental data. Welcome to the show.

Pleasure to be here. Thanks for having me in.

Yeah, so tell me a little bit about kind of your path and how you've gotten here in getting to data science and the environment and everything that you're doing now. You've worked in quite a few different areas to get here, so share a little bit about your journey on how you got here.

Well, it's certainly been a long journey, but I mean, I was always focused on sustainability, climate change, biodiversity, these, these large questions that are sort of classed as these existential risks to, to our place on the planet and our continued survival. Um, and I always wanted to work in this field. Obviously, I, I grew up in, in the '90s and, um, early 2000s, and, and climate change was very much on the agenda, uh, throughout my education. And I just always felt that was something I, I would be working on, and I deeply wanted to work on. So, um, I studied essentially physical, physical sciences, so, um, physical geography really, but it was earth sciences and earth systems. Um, I studied, uh, up to an undergrad level and then, then, uh, went out to, to work in, uh, the implementation of the Kyoto Protocol in, in terms of like carbon trading as, as a carbon analyst for one of the big banks. I quickly felt like that Despite the marketing, that wasn't really going to live up to the hype of making a significant difference. So I thought again and went back and did a PhD again in physical geography, in particular climate change. And then I spent about 8 years in fundamental research in the public sector across a PhD and a number of postdocs traveling around the world working on projects around clouds, aerosols, remote sensing, the impact of climate change on different types of ecosystems. I published a lot over that time and had a lot of high points in my career where I got to work with groups like the IPCC and was a contributing author in the 5th Assessment Report. And even though I was having good success and getting traction in my very narrow field, I felt like it still wasn't really doing enough. Having more precision in our understanding of the impacts of climate change isn't very helpful right now. We, we're well past the point where we need to take action, and we need that action to be at scale. So I became quite frustrated at my ability to kind of affect change from that position of science. So I gradually kind of moved into— over the course of my research career, I moved into software engineering, particularly with respect to how science gets done. Like, how do we actually how are we confident in our results? How can we reproduce our results and be sure that we really know what we know? And that led me into more the technology side, uh, an engineering side. So I effectively transitioned from academia into research software engineering, and then from there into ESG-related tech consulting. So I worked for quite a few years in essentially data science or applied science, working at this intersectional point between technology, product, and domain science at the core of quite significant projects for NGOs, private business, and governments. Projects like Global Forest Watch, Half-Earth, Climate Watch, Resource Watch, Trace, Aqueduct, these kind of large-scale projects for monitoring resources on demand that very much fall into the— under the umbrella of ESG. And that really helped me develop a lot of skills beyond just purely the scientific into how is this being applied across the entire global marketplace, everything from the sustainability side to impact and the economics, but also very much product-focused. So how can you ensure that you design a system that people can actually use that condenses down complex datasets into actionable insights in near real-time at scale.

Always a challenge.

Yeah. And from there, it was a short step towards Survest where I'm head of innovation. So, I'm essentially sitting formally at this intersectional point now between science, technology, and product. and where we serve essentially a zero-to-one function in the company, the group that I'm leading. So we build out new key competencies and integrate them into our products. And all of these products are around global-scale sustainability, understanding both adaptation and mitigation with a particular focus on the physical risks of climate change down to an asset scale. So we're really trying to give everyone the tools to understand how is climate change going to affect you down to your single site level, and how can you plan, make decisions that take climate risk into account.

Right. That's a long journey to get here. But it sounds—

I think about 15 years.

But it sounds like it's a really good path in order to get to sitting in this intersection You know, you had to have the experience of working in the, in the raw data side and understanding what that is. And it's great to see that you guys are tackling the challenge of how do you take— because we've had raw data, right? We've had all kinds of scientific data for decades and decades about the environment, about all kinds of stuff. And, you know, it's been sitting there, right? The average person Doesn't even look at it because it's so—

It's not decision useful. Yeah, exactly.

You can't look at it and even guess, you know, it's like a 3rd grader looking at calculus and then asking them to solve it. It's just not— it doesn't connect to that usable level. So, and you mentioned down to the asset level, right? Down to a plot of land.

Mm-hmm.

And the impacts of that. And so what kind of— what— how did you— how did you guys arrive at being able to— what's that kind of level of detail? How does that— you know, when you look at your impacts and, and you said like physical damage, like being able to see where that actual damage— is this more predictive? Like, are you, are you looking and modeling it out and saying, here's, you know, over the next 5 years? Or is it more focused on, you know, how, how are you looking at that?

Yeah, so, so you're completely right in your diagnosis that in fact we've had these data for a long time in crude coarse forms, uh, but they've been siloed across different domains of, uh, domains of remote sensing, of weather forecasting, of climate modeling, seasonal predictions, of historical integration of data to create what we call reanalysis projects. We've had access to these data, but they've been in spatiotemporal regimes that aren't readily comparable, and they haven't— we haven't been able to integrate and get a view of risk across time and space that's standardized, that's multivariate, that— that really gets us down to decision-grade level of understanding. And that's been a key problem. The science obviously has been available for policymakers for many years since IPCC has been publishing the first report. But obviously that hasn't really been acted upon very effectively yet. So we're inverting this and we're taking it down to asset scale and distributing and democratizing that intelligence so that everyone can act on that. Now, the question— sorry, your question about 5-year timescales and how it works down to asset granularity. So it's taken about 5 years of research to get to this point. And at our core, we have what we call a proprietary Bayesian signal framework. And what this really means is that we use a lot of quite advanced statistics to harmonize these data from disparate sources and bias correct and downscale it so that we can understand both probabilistically what are the conditions across all of these sources of truth harmonized and unified and standardized together, and, and also enhanced spatially to high resolutions and bias corrected so that we can really kind of have a, a high enough level of confidence in the data to be able to make statements about the probabilistic conditions at any location. for any time period of interest. And, and the time periods that we're serving right now, they do range from 1970 out to 2100 across 3 different future, future scenarios. These are the scenarios that, um, essentially are used to understand how might the future be by groups like the IPCC. These are emission scenario pathways, uh, fundamentally. So what different types of, of climate forcings are we, are we going to be experiencing? Will it be a best case where everyone pulls together and really gets on top of the mitigation, the decarbonizing their assets and puts us on a Paris-aligned 1.5-degree world where the climate risks are there but manageable? Or are we going to be more on the other extreme of a business-as-usual world where we just keep emitting as if there isn't really a problem? keep using all of those fossil fuels and end up with quite a destabilizing degree of climate change that has extreme consequences. And if anyone wants to understand more about those, I'd recommend The Uninhabitable Earth by David Wallace-Wells to get a primer on that. But yeah, we really have this range of possible futures, and we look across all of those futures and let our users understand what are the risks to those assets across those potential scenarios so they can plan as they see fit for the type of risk that they're willing to tolerate.

Right.

And we also integrate a range of different types of views of the future, everything from long-range weather forecasts to seasonal forecasts to what we call initialized model runs. So model runs that start with the current conditions and then look forward a few years to try and capture the El Niño, La Niña phenomena as well as the long-running climate models that are used by the IPCC. So we bring all of that knowledge together, so the timescales come out of integrating that information. And we really have a range of timescales because the things that our users want to do with that data varies, everything from operational planning and monitoring to decisions over credit timescales of 36 months to more like merger and acquisition planning or P&L forecasting, whatever their needs are, that ends up determining what time horizon they're interested in looking at. So, we abstract all of that technical and scientific complexity away from them and give it to them in an open network where everyone can see and share the same understanding of risk, which is pretty critical because right now we are in a world where people have access to this information. As you say, it exists in many different forms around the world. but often in ways that aren't useful for making decisions at a granular scale. And the people who do have access to that data in a way that they can make some decisions are people who can pay for it. So they're the—

Right.

Let's say it's information asymmetry. Well, actually, government wish they had access to that. There are use cases that come out of combining the data in the way that we're doing, which are particularly interesting for the government, and that's going to be a critical use case going forward, understanding things like stress testing on national infrastructure and the security implications that come from that.

Right.

We still have to work towards the world where we have that level of understanding, but fundamentally, this understanding needs to be shared. So everyone needs to understand risk in the same way to be able to have a conversation on a level playing field. We're taking the game theory out of deciding what to do about climate change with respect to adaptation and mitigation. So it can't be, it can't be I win and you lose. It has to be a win-win. We all lose if too many people are on the losing side of climate change at an infrastructure level. If we just end up with large components of our, the system that our economy depends on collapsing due to stress that they're not prepared to handle or risk that they're not resistant to or resilient to. then we all lose. I mean, and we're seeing a little bit of that at the moment in terms of supply chain issues that are affecting us globally. When you have people in London worrying about getting their PS5 because of chip shortages, it's, you know, it's starting to see this kind of integrated issue, the issue of having such an integrated, interdependent system that has high sensitivity to shocks and stresses that is not currently resilient to, and that, that will only exacerbate over time.

Yeah, that is one thing I would say that, uh, 2020 kind of woke up a lot of consumers out of. We are actually— the, the idea that, um, we aren't a global economy is dead. It died a while ago, but people finally clicked in that it is very integrated when you When you have a massive problem in one country on the other side of the planet, you're impacted in some way. Today it's, yeah, you can't get your PS5. So that's not the worst thing in the world. It's not damaging, but that's indicative of what can happen should the wrong supply chain get damaged. So wait, You guys are doing— you mentioned there's a lot of data that you're pulling together from a lot of different sources to be able to put it together. And you mentioned, you know, asset management and being able to look at, you know, the risk associated with a certain area. And you also mentioned supply chains. So are you able to start mapping together basically the supply chain? I almost feel like kind of what you're building is a Global resource inventory system. Like, you know, if you're playing a game, you could, you could look at a planet and know all your resources. I feel like that's kind of what you guys are doing is you're, you're pulling together and saying, hey, look, we're all using all of these resources. Here's where they're at. Here's, here's the ones that might go away. Here's the ones that could, can make it. So is supply chain playing into that? in what companies are wanting to build, build out, be able to kind of future-proof supply chains?

Yeah, so initially our focus is on understanding built infrastructure. So those are freestanding permanent buildings, basically all types of buildings. We're also adding to that different infrastructure types like linear infrastructure roads, rails, pipes, power grids, as, as well as looking into the future natural capital assets. So, um, everything from sourcing regions to agricultural holdings to green spaces. Um, once we've integrated all those asset types, which, uh, we will do over the course of this year, then we'll have a complete view, as you say, uh, almost like a video game of the world. And we'll be able to, um, we'll be able to determine things like these advanced stress testing capabilities, network risks, second-order risks. Um, yeah, I mean, at the moment our focus is on, is on buildings, uh, and the damage and replacement costs associated, uh, with climate hazards impacting those buildings. Um, but we, we have to build up, uh, this archive of the world's built infrastructure, uh, into this open network so that everyone sees the same thing. Uh, and we fundamentally use the, the characteristics that we're extracting on all of those built assets to drive our platform capabilities like search and discovery and segmentation, um, and the ability to group these, these different types of assets together into indexes of risk across different sectors, industrial sectors or regions. Um, so I mean, the simple answer is, is yes, but there's complexity because obviously—

Right.

If you want to isolate a specific supply chain, Then there may be some extra work needed to be done by the users to go in and identify these are, these are my suppliers' assets, or this is, this is my sourcing area, or this is a critical supply line. But once they, once they build that out, at first in a fairly manual way, but then later in a much more automatic way, then they can have this holistic view of risk that they have both directly and indirectly. And that's really revolutionary, revolutionary capability. There are a lot of players right now in the market working on carbon. Carbon is sort of the low-hanging fruit, and everyone is very much over-indexed to worrying about the carbon footprint and the mitigation side of the problem. But it's really only half of the story. We have to also worry about adaptation. Fundamentally, the amount of emissions that we've already put out collectively, go humanity, has committed our planet to a degree of warming that is already going to be highly problematic to deal with.

Mm-hmm.

So whether or not we magically solve mitigation, we magically decarbonize our assets, we'll still have to deal with inherently high levels of physical risk. We need to make sure that our assets and our infrastructure are resilient to those risks because it's almost an equilibrium problem. we built all of our infrastructure considering climate from, say, last century that was very stable, very consistent, very predictable.

Right.

But we've shifted our climate from that static baseline into essentially an unstable condition that's changing year on year. This shift is basically causing our infrastructure to be what geographers would geographers would term, um, relic landscape. It's, it's out of equilibrium with the current environment. It's not fit to tolerate the conditions that it will be exposed to. So this is really the other side of the coin that Sylvester, uh, and our technology deals with tackling, and it's quite unique. There's not many players in that space because of how hard it is to actually tackle this problem. There's something like 90 different carbon players and only 3 or 4 different physical risk players, but Once you, once you crack physical risk and unify that view of carbon, it becomes incredibly potent because you're able to then make joint decision-making for the first time. You're able to look at your collection of, collection of assets and say, uh, not just drive, not just decisions on how, how you want to invest money to adapt your resources, but also to mitigate carbon. So it's basically joint decision-making between adaptation mitigation, which is really what we need. in this open, transparent, standardized network.

So with that, do you see kind of the data that you're pulling together then— well, obviously there's the carbon data, right? And that's kind of, as you said, there's lots of people. There's, I think Elon Musk has a new X Prize all about carbon. Microsoft's invested. Like, it's very much a public buzzword. You can Reduce carbon on your shipping by clicking a checkbox when you order. Like, it's very, it's very in front of everybody and kind of, you know, if, if it keeps going on its path, right, it will be addressed, carbon. But you're pointing out this other portion. And so are you seeing this as something that companies kind of— do you see this as something that companies are saying, hey, we need this? Or is it just a blindside? That there's not really anybody going, we need this because we're aware of it. Everyone's just not— we're just not aware of it. Kind of how's that dynamic for you guys in that half of this equation?

Yeah, so we work with some of the biggest companies in the world, and no matter their market cap, they're all really in the same position, which is the state of climate intelligence infancy, they, they don't yet know what they don't know because they've never been really keeping records of how conditions have affected their operations. Um, so one of the, one of the first things that surprises them is the historic baselining. So when they come into our system, they're able to get this view of how, how has— how have the climate, the properties of the climate system been over their, their past activity from 1970 to present day, as well as looking out into the future. So this baselining and forward projection really kind of set the stage for them having this initial, initial realization of the, of the benefits of climate intelligence. Um, but it goes far deeper. And one of the critical use cases that, that they have to engage with over the near term, uh, uh, uh, from, uh, disclosure and reporting use cases. So there's regulatory pressure for, for companies over certain sizes in certain regions to start to disclose their physical risk, their material risk on the assets that make up their portfolios. Yeah. And that's impacting a lot of people. I think 21,000 companies will fall under disclosure laws over the next year.

Wow. That's not a small number. I was thinking less. 21,000 is a lot of big companies too.

Yeah. Yeah. And almost no one really has an idea of how to do this. They're just starting out, and it's quite a shock to them to grapple with this problem. So, one of the key initial products that we're putting out is an automatic reporting capability. So, TCFD-aligned automated reporting that will be standardized, transparent, explainable, version-controlled, citable. enable the users to download the components. Everything's been produced with teams of data scientists, visual designers, engineers who are basically calibrating all these technical analyses to be understandable and extractable elements that they can put into their own market-facing reports or just simply publish via our system and have their— have them live online being these transparent disclosure reports.

And with this data, right, they're seeing this, is there— is it kind of an aha moment at all of like, oh, we've figured out why we're not doing it? Like, is there some tie-in to business where you're able to correlate, oh, look, we've looked at it, we've now figured out based on the data you're able to pull together, We should have been doing better. Like our assets, we have higher risk. We've actually had losses. Are you able to start tying that into a direct, you know, financial model to say, yeah, look at look at the impact that the last 1970 has been 52 years, right? 52 years. This is what's happened over that course. And are you starting to be able to pull that together to show? show companies like that exactly what that impact actually starts to cost in dollars?

Uh, so you're— is that's exactly the right question to ask. Um, the first product we've announced is EarthScan, which is the enterprise-facing product that I was just explaining that will let you get this baseline of risk, the future projections, share it internally and externally, and create your automated reporting. What you're just explaining is, is the next logical step, but we have not yet officially announced that product, so I don't think I can go into detail about it. But let me just say that you're, you're on the money in terms of your, your suspicions, and it's, it's going to be a critical piece in solving climate change, as capital markets are, are going to be the main levers for driving adaptation and mitigation at scale across the global marketplace. We absolutely need their buy-in. We need to reprice transactions on credit and insurance decisions, and we need that as fast as possible because this is really a problem that we can't afford to delay anymore. We've spent 40 years of delay that we really should not have had. The science has been clear for a very long time, and we have not acted.

This is— I feel like this is the answer I was talking to someone else about. Doing a true cost comparison between like solar and coal power. And they brought up basically we're in a carbon subsidized market, right? We look at the cost of fuel, but we don't actually look at the real cost of using a gallon of gas. Whereas, you know, building a solar panel, there's nothing to— nobody's hiding anything. There's no— there's less to it. But we're also still not looking at the carbon impact of that. And so we just had this conversation over if you actually were to put real cost to real cost, including these impacts.

Now the externalities.

Now all of a sudden you go, well, of course this option is cheaper. If you actually look at the cost. So I think that, that kind of true, like you said, an unbiased, removing any biases and actually neutralizing that data will allow people to actually do an apple to apple instead of, you know, I think we're doing a zebra to apple right now. We're like not even the same group of anything when we do a comparison. So that's great. That's great.

That's So you're completely right that the risks of climate change are shared, and that externality problem has been a significant problem in really understanding the true cost of climate change. And that's exactly what Sylvester is going to enable people to do, to price in those externalities into their operational planning, their P&L forecasting, their long-term thinking for the first time.

That's great.

That'll be—

and it sounds extremely useful. If you guys ever want to dive down all the way to a per-part level, I'm sure that would be an interesting comparison to dive into supply chains.

Yeah.

And true carbon and true impacts. Like, that really— that's what you're doing is you're measuring true impact of what a company is doing on a, you know, detailed but still larger scale, right? You aren't measuring each screw, but you're saying this is what you're doing as a company as a whole. This is the impact that's happening.

Yeah, exactly. Initially, it's an understanding of damage and replacement costs and disruption to business operations. with respect to the physical built environment. But as you say, there's a lot of pathways to develop this technology along, and there's a lot more to do past that initial phase of capabilities. And we will be keeping a close eye on developing those as our—

As our—

As our company grows, should we say.

Right. Yes. Well, I'm sure those are all resource intensive, so no one's going to expect you to pop them all out in the next year. That would, that would be crazy to do.

Yeah. There's also, there's also a learning journey that users have to be taken on, right? Because at the moment, markets are very unsophisticated. They, they, I mean, ESG was essentially a failure. It didn't really help us reprice anything. It just gave very subjective metrics that could have their main application in greenwashing, to be honest. And, uh, what we need to get to is a standardized, uh, quantitative system that lets us price in externalities into our understanding of transactions. Uh, and yeah, that's exactly where we're taking this. So it's— say it's going to just blow ESG away and become a new standard, but people have to People have to get— have some time to learn what this thing called climate intelligence is. I mean, the market's already fairly comfortable with carbon accounting, emissions profiling, target setting for carbon, but they really are infants when it comes to understanding physical risk. And what we're bringing to them is this complete picture of physical risk with carbon. and pricing. So it's, it's everything coming together, and it's going to take us a while to, to build our products in a way that, that take users on this understanding, that, that close this cognitive gap for them. And we really put a lot of effort into, into making sure that the products are designed in a way that, that takes out a lot of, a lot of the stress for our users and the strain of understanding how to go from insight to action.

Right. And, and when you're— so a lot of this data, right, it also is pointing out to companies and even governments, obviously, they would want to be looking at this as well. You know, it starts to point out like, hey, you have, you know, whatever, these bridges are going to fail at this point, or this building isn't— is going to be too risky, the risk is just going to be too high, which I'm sure at some point an insurance company will go, We're not going to cover you because it's just, it's just too risky. So, yes.

As many people in California are just discovering. So, so we have, we have, uh, insurance companies use, use single risk models. They do understand this type of risk, but they just don't understand it in a way that is, uh, it is, um, unbound from the historical distribution. The insurance companies usually Look at past claims data to try and understand the risk in the region. But that fundamentally doesn't work for climate change because the past is not a reflection of what's to come.

So it's weighted. It's just, it's a weighted model on past more than trying to look at what we know is going to happen.

Yeah, exactly. And what we provide fundamentally are probabilistic signals into the future in terms of how frequent things happen, how intensely they happen, and how long they last across multiple variables of risk. So that's the exposure component of the climate change, which acts as a threat multiplier, basically.

Right. And so with all this data, you know, you're kind of— you're being able to, you know, point into the future a little bit and identify those. Obviously, our, you know, our construction methods are based on what worked in 1900. So we're— I mean, we haven't really updated much, um, and some of it's older, right? We— how we build things— I was talking to an architect, and, and how they design and build buildings just hasn't really— it evolves very slowly, way far slower than what the environment is changing. So is there, is there this intersection where there's interest or— where this can start to be applied into creating solutions and saying, well, this is how we build a building now, but here's how you should build a building in order to make it to 2100 and not have it collapse and still be functional. Are you, are you looking to or already having those conversations of directing this into modeling what we need to build down the road?

Yes.

How do we change the regulations for, you know, build construction code?

Well, we're not looking at the regulation side of construction code, but, but, um, we're most definitely looking at partnerships, um, around civic and architectural engineering. And, and what you said was, was completely, was completely hitting the nail on the head there. Um, The problem is that buildings are essentially built with this concept of return periods in mind. So, for a given location, you'd want to understand historically how frequently did events that could cause damage or disruption to the building occur. So, for instance, if you're near ocean or a river, you'd ask, what's the 1 in 50 or 1 in 100-year flood levels likely to be from this location? And then you just make sure that you've designed your, your building to withstand either a 50-year return period or 100-year return period, depending on, uh, the quality you wish to build to. And obviously there's a trade-off there in terms of how you're spending your money to build. And in fact, that trade-off, that trade-off has other implications as well, because when, when you look at the developing world, the developing world often prioritizes, um, low cost over— Quality. Hardening their infrastructure or having it, having it even be built to withstand, uh, what they understood the historical return periods to be. And that whole thing goes out the window when you consider climate change, because what climate change does is completely mess with that, that understanding of what is a normal return period. Like, what kind of depth of flood would you expect once every 100 years? That completely changes under climate change. That's why you'll see people talking about in the weather, more extreme events, more frequently, more intense. Like, this is basically the crux of the problem. The extreme events are coming more frequently, more intensely. They're lasting longer. They're coming at different times in different locations. The patterns of weather variability are changing whilst you have this background of the climate crisis with essentially chronic conditions. So for instance, like sea level rising. the storm surges get higher and higher every year. Um, the, yeah, the rainfall patterns can get more intense, uh, due to climate change, which causes like increased, uh, flash flooding or pluvial flooding. Um, all of these factors combine, uh, and you need to consider them all together to really get this holistic view. Um, and there's one other component there, which is the asset understanding. You have to really understand what, what is this asset, what, what is it, uh, Functionally and materially so that you can know whether or not you need to worry about the changing conditions. An easy way to think of that is the analogy of the 3 Little Pigs. So in the 3 Little Pigs, obviously you've got the big bad wolf blowing wind is essentially a risk of wind gust. So the wind— the wolf is a wind gust forcing. He's giving the same forcing to 3 different types of assets, the sticks, I don't remember the story properly, but something about sticks and stones and bricks or something. But assets of 3 different types, and 2 of those assets are sensitive to that initial wind gust forcing, whereas the 3rd one is hardened. It's made of bricks, so the wolf can't blow it down. But that's essentially what you're dealing with. You have to understand the material properties of the asset, classify those to really understand, like, Are these— are these, um, is this exposure risk, does it matter to this asset or just, or to this chain of infrastructure, the supply chain, or this, um, infrastructural network that we rely on for transport, um, or whatever it may be?

Right. And that, that's a lot of classification. So yeah, that's a, that's a lot of— is there, and is, is there a manual part to this where You know, you basically, you're working with companies and they're, you know, they're going out and looking, having to look at their building to see how it is. Like, how, like, are you doing that kind of detailed work? And is that more manual? I mean, is that— because I feel like that's not— there's not really a good construction database of the world that exists.

Yeah, you're completely right. This is— there isn't one. There isn't a database that we can just go to and just extract and have all the answers. You're completely right. So building this up is part of our proprietary IP. We enable open access to the network so you can come in, select an asset, and then understand the risk, the exposure risk, and the impact of that asset. But the process that we have of actually populating that asset data, of labeling it so that we can characterize it and segment and do all of the clever stuff that we need to do on the backend. That's very much where the secret sauce of the company is. We basically have 2 strands. We have characterizing the world's assets and we have characterizing the exposure risk across time and space. So the asset work stream takes significant investment. Building out that dataset, it will be a unique one-of-a-kind dataset when we're done. At the moment, it's We've got good coverage across Europe and America, and we're building out as we're going.

Right. Well, this is amazing. There's a lot of— you guys are being able to produce a lot of useful data with this, which is great. It's been decades. Now we're finally— finally somebody decided to make it usable, which I think is—

Well, it's taken It's taken exponential technology development to get to this point, right? So as you said, that we've had some of these data for a while, but we really haven't been able to leverage the, the mega-scale computing. We haven't been able to leverage the, the high-fidelity location data, and we haven't had the, the ML, the AI, the computer vision that we've needed to get to this point. And it's, it's really only become possible over the last few years of research. So we spent about 5 years of deep R&D, particularly around developing our own signal processing framework and the workflows around assets. And it's really only now just becoming possible to achieve this.

That's amazing.

It's an exciting time. I mean, from someone as a climate scientist and someone who cares deeply about what happens to humanity as the century unfolds. I wish we would have had this sooner, but it's simply that the time is now for this technology to rise and for this category to be built.

Yeah, this environmental intelligence, that's a great, great way to put it. We need more of it.

Exactly.

We need it in our businesses in order to make decisions.

Well, the category that we're building is specifically climate intelligence, but it's under the umbrella of climate tech. But it's a brand new category that effectively hasn't existed before in the way we're bringing it to market.

That's awesome. Well, I really appreciate you taking time to talk about what you guys are doing, and I think you've done a great job of explaining what you guys do. And it's amazing to hear, hear it, just that you're, you're able to even do this, right? It's an accomplishment just to be able to say, we can make the data useful, and, and you're doing it. So it's really quick, if anybody wants to get ahold of you guys and start working with you, what's, what's the best way to do that?

So, if you want to fire any comments or questions at me, I'm on Twitter @BenLakin. I'm also fairly discoverable from internet searches. You can find me on LinkedIn as well.

We'll make sure those are in the description, so that'll be easy.

Thank you.

Fantastic. Well, thank you so much for being on. I appreciate it. It's been great.

My pleasure.