 products we have, everything is life-only specific stuff. Feel free to add me here, probably. And I was really small team in Montreal. There are six people. It's actually a Scottish company. We're based in Edinburgh. I have three developers, myself included, all in Montreal. And some of the people, they're all analysts. These people are different masters in either geospatial stuff or environmental things. Basically, they're all that kind of profile. They're mostly scientists. So what do we do? Basically, what we do is, do you like a little fish? I see you. So what we do, we do things about, mostly about climate change. And we measure both ends of climate change. So we try to measure things that are making climate change worse. And then we're trying to measure the actual impact. And we have two main products to do that. We have arnipax and we have an arnipereal system. They're two very different products. But you'll see that they're extremely related. They're just completely different technologies. And they're completely different things. Arnipax is something that's used for something that's called carbon accounting. So what is carbon accounting? One example of carbon accounting is if you go to a grocery store and you pick up products. And the product has a little green icon on it. And it says carbon neutral. What the hell does that mean? What it means is that in the course of making any product at all or conducting any business, you've got people on the road. If you're making products with the machines, you're probably to see by natural gas. You're burning that natural gas. You're buying materials. In the course of doing all of that, you're creating emissions. And the emissions can be CO2. But it can be also all sorts of different other gases, like CH4 methane. You'll have refrigerant gases, which are really, really bad, of global warming. What these companies do is that, A, they'll try to minimize that. And B, to be carbon neutral, what they have to do is that they have to buy carbon offsets. So it means that if at the end of the year, they've measured that they've committed, for example, 100,000 tons of CO2, what they'll have to do is pay someone, and that someone is going to do something like plant a lot of trees in Brazil, so that the trees will actually capture that carbon back. It's not exactly the same carbon, but the idea is that overall, in their activities, they're carbon neutral. But that's kind of hard to do. A, how do you know how much you need to buy? How do you know how much you need it during the year? Of course, your activities, all you know is you bought 3,000 tons of this sort of plastic, you bought this many millions of cubic feet in actual cast, and you have people on the road, and they have these cars, and they've driven for 100,000 kilometers. But that's all you know. It's really hard to get from this point to the actual number of emissions that you've released in the atmosphere as a carbon activities. And then on the other hand, you're paying this guy for really good offsets, and he's telling you, yeah, I'm giving you money, and be sure that for that money, I'm going to capture 100,000 tons of carbon. How do you know that's real? You don't really know that. So our ecosystem is on that end, so our impact is about knowing how much you've made it. Our system is about certifying that what he's saying is actually real, and you'll see how we do that. I'll go in the tech stack right after that. So if we go to our impacts, our impacts, it's carbon accounting, I mean it's got a crowning in it, it's going to be a really boring product. So that's what it looks like. That's one of the pages. It's not an open side, this is something that is, the companies need to pay to do that. But what it basically is, is that the companies will be, yeah, let me back up a little bit. There's two ways of doing this, and traditionally what people have been doing is that they have these huge spreadsheets, and they'll enter, for example, the number of kilometers driven by their personnel in the company owned cars, and then they'll dig around the web somewhere, and they'll figure out that, well, these are kind of their Ford Focus, let's say they all have Ford Focus, it's real simple, and they'll try to figure out, by getting the numbers, like the efficiency of the vehicle, and then how much carbon per liter of gas it's going to emit for that specific car, and they're going to get a multiplier, and then they're going to multiply like that. Okay, I'm done with my cars, and that was a whole day of searching for data. But then he's got everything else about the company, there's a lot of stuff. If they've got offices in 10 different cities buying electricity, and buying natural gas for heat, it's all going to be different multipliers, what we call factors. What our impact does is that it automates everything. We have tens of thousands of those factors, and the way it works is that you'll be able to hear what you have is a company that's spouts aluminum in Kazakhstan. So there's kind of a set up phase where we ask questions about what the company does, and what kind of things need to be taken into account. But once that's done, you'll have a question here, for example, which is the annual consumption, when you actually, when you make aluminum, have anodes got those, these things actually burn now, because that's how it works, and you need to replace them. Well, the system is able to know which units you can answer the question in so that it will be able to actually get the emissions in the end. So I'll be able to do kilograms here, just making sure I'm not on the demo site and not changing the live data, I am. So it auto saves too, so. And you answer a bunch of questions like this. If we go back, I'll go back, but if we go back to the car question, how many kilometers you've driven, well, there's going to be a section in there that says, well, for the New York office, what was the company old car activity? It won't necessarily just be kilometers, the companies don't necessarily have that data, they might know exactly how many kilometers were driven in what type of cars, or they might actually not even know that, and all they know is how many liters were consumed. Or not even, all they know is that in April of last year they paid $20,000 in gas. Well, what you need to do is $20,000 of gas in April last year in New York, how much gas did that get you when it got through this much? And for this time of car, with the liters, how much CH4 would be emitted? That's the kind of chain of calculations that it's going to do, and it figures all those things out based on the time frame, based on locality, based on the activity, and it figures out what units it can take to get to those gases, and then it adds everything up. What you get in the end, something like this. So I won't get into the carbon accounting, there's no scope things, it's really boring, but basically in the end, that's the kind of stuff you get. These are the tons of CO2 that were emitted per type of activity. So premises is things like the electric unit for the different things, for the different buildings. Business travel, that will include even things like a hotel stay in Helsinki in March last year. Well, on average, you don't know precisely because you don't know how much water he actually used in his car, maybe this guy spends an hour and a half in the morning. You have no idea, but there's average data, and that's a big part of what R&Lists actually do, they just, they mine for data around the world. It's really hard to get these kinds of factors in China, for example. But then they mine for that, and they import it to OI, so that we don't do that. But it's a lot of data, the calculations engine is actually pretty neat, and actually Michael Mully worked on that a lot, and he did most of it. But it gets fairly complicated. Another example is if your building is heated by electricity, how much, what are the equivalent emissions? In some cases, all they'll know is the electricity bill is month by month. So what we have to, what the system needs to know is that how much did $300 get you in electricity in Jevexity in February last year? The electricity prices per month, per month, per locality around the world. But then where does electricity come from? Did it come from coal, from nuclear? We have those things as well for pretty much the entire world. So depending on how the electricity was generated in that area, it's going to result in very, very different emissions. So for that company, you end up with, here's an example of the factors, what I was talking about. So basically, I'm looking at an office in London and we're looking at the travel of their employees, my trains. So what they've done is that they've been able to enter how many passenger miles have been used. So this is if they had them people traveling each 1,000 kilometers, then that's 10,000 kilometers. So they have passenger miles or convert to kilometers and basically the calculations engine figures out all the steam calculation, automatically for that place, for that time. And then it knows it's got a factor for national trains that are intercity in the UK in that time. And we know that it emits 0.05 kilograms of CO2 for maximum nuclear and it has that for CH4 as well. It has that 0.20. So overall, you end up with a number for each type of gas. We also have uncertainty associated with all those things. And the other thing that is kind of interesting is that CO2 is the one we always care about, but the other gases actually factor in, they'll have different global warming potentials. So there's a multiplier that's where you take the final amount, that's the final amount of CH4 associated with that activity, but there's a 25 times multiplier, which means that a one kilogram of CH4 has 25 times more global warming potential than CO2. It gets to 298, but actually with things like refrigerant, yes, it gets into thousands. So this is the product, this OIA is how the companies measure their total emissions for a given year. And that's how they get that final magic number so that they know how much offset is applied. Hawaii is all built on top of, it's all on Django, it runs on Ubuntu, on Amazon EC2, most of us databases. So it's all using all those projects, I might as well mention here that we're on GitHub. And we'll release more stuff. There's a lot of internal technology that we do want to release, especially social network ecosystem, which I'll get to in a minute, it's more interesting than I want. But there's a bunch of things in here that you might find useful. If Negra's kind of neat, it's a few lines with my Michael Mali, it lets you do translations in any language of your data, but using PO files. The other projects basically add fields in your database. This one uses PO files so that you use the normal get text facilities. So we've measured how much carbon we've made. We went to the market, and we bought emissions offsets from some kind of shaved street. Is it really growing those trees, or is it parking money and pretending to? We don't know that. Well, yes, we do. What we do is that our ecosystem is the other big product, we have PCT Biocarbon Tracker, that's one, it's kind of an instance of it. The way OE works is that we'll have different instances of the same product for different companies, and they'll basically be paying us to create an incarnation of the product. So Biocarbon Tracker was done for Greenergy, which is the largest seller of biofuels in Europe. Why do they want to do that? The reason is that in Europe, those regulations coming online, this is where the politics come in. If you want to sell biofuels in Europe, you need to prove that whoever you're buying, you're buying the sugar cane, or whatever else the corn from, hasn't been deforestation in large areas in Brazil, for example, or you didn't take a huge march that was huge in my own adversity with 100 different species, and they didn't just fill that up so that they could grow sugar cane, so that they could buy, they could sell Greenergy the sugar cane, who would make biofuels to sell in Europe, for a little hippies who thought they were doing good, but they're destroying forests. You don't want that. So those regulations coming online, the problem is that there's no tools to prove that. There's no certification yet. So that's what we did. And what you see here is part of Brazil. You see some dots which are, oh, there we go. So that's one of the farms that is selling sugar cane. Yeah, like in Brazil, there's not really any conversion to biofuna, but actually there is, there's lots of big mills, so that they'll actually convert the sugar cane to biofuna and ship that back to Europe. So these are the places that Greenergy buys from, but Greenergy doesn't own those places and they wanna make sure that they won't be a backlash two years from now when journalists find out that all the sugar cane they've been buying has been helping deforestation in Brazil. What you're looking at now is the carbon change between 2005 and 2009. It's something that as I said, we have team scientists, they've come up with algorithms to do that. This is the only map in the world of that type and we have that for an entire globe at 100 meter resolution. What you see towards the yellow hand is carbon is places where the carbon content above ground has been decreasing. The green and blue carbon content has been increasing. So overall it's not that bad in Brazil. There's that stuff happening, but overall it's getting better and better. But this way what is going to happen is that Greenergy is going to look at the specific outlines of the farms and they'll be able to know whether the farmer has been deforestation around this land or things like that and if that happens they just come off and that's it. Here is another layer This is basically the carbon content above ground. So we have special algorithms where we take images from salamites and we're able to deduce the carbon content. I have no time really to talk about it, but I'm happy. These algorithms are actually open through the scientific articles about them, so I wouldn't mind talking about it. Just a little bit of time here. Carbon risk index is kind of cool. What we do is that we combine the carbon data with other data and what we do is that we're able to tell you the red areas are places where we think that in the next couple of years there's a lot of carbon at risk of being lost and released into the atmosphere. So those are the places that should be watched. And here's a new one, deforestation. This is actually brand spanking new and we don't have that much data yet. Oh, here we go. I can barely see it. We came up with deforestation algorithms. What we do is that we take data from a certain satellite every 16 days. We're able to tell you, we're able to monitor regions very, very specifically and we'll be able to send you a report saying that part was just deforested now. So all the people buying those things from the forest and wanna make sure that there's no deforestation or if governments create parts and wanna make sure that it's not degrading, they can get alerts when that happens. And this is, again, something that goes around the globe. To give you an idea of scale, right now what you've been looking at is 30 arc second resolution, which at the equator is around 900 meters. That sounds pretty big, but it's for the entire earth. If you look at the map of carbon, that's here. If you look at the map of carbon, it's going to be fresh eventually. There are, in the database, we have a row for every single pixel, that's about 900 meters square, and that's 222 million rows. So when we started this, we were looking at small areas and we were thinking, no problem, Post-GIS has all sorts of things for these things and it's awesome, you can tell it. Give me the rows whose coordinates fall within the public's polygon, it's easy, George, I'll just start. It works for small areas. When you get to the entire world, the query times were around 15 minutes. But the problem is that I want to be able to do this. Let me just zoom. I created a small polygon, any arbitrary polygon. It's telling me how many tons of carbon above ground there are, and you notice that took one or two seconds. So that was one or two seconds to query three tables, each having around 20 million rows. I have the amount of carbon per vegetation type because that's another thing we do, we're able to determine what the type of vegetation is for every cell. Is it forest and what type of forest is it? Is it grassland, is it desert, is it water? So this table queries all of that and it's able to tell you how much carbon there is per vegetation type. And again, there's a risk map here that tells you how much of that carbon is at risk right now. Being able to do that, that's part of our core IP, however, that's one part I'm not really able to do to talk to you about, but being able to go from having lat-long coordinates and doing queries that were running for 15 minutes to a system where I can do these queries within one or two seconds, query hundreds of millions of rows. That was pretty big and that's what's allowing us to do that. The rest, in terms of actually showing you the tiles, that's all using open technologies, most of them in some applications, but that's pretty much it, there's a lot of caching, obviously, and a couple of terabytes of data, but it's nothing that hasn't been done before. So I'm out of time. I'm not quite over yet because I'm sure that all of our speakers and most of you think they are probably thirsty. So we're going to head out to Minidux and we'll get two free beers, every one of us. If you get the free beer, you have to follow Mathieu. Mathieu is going to be the free beer coupon. And that's it, we meet you there.