 Hello, and welcome to theCUBE's presentation of the AWS Startup Showcase. Data as Code is the theme. This is season two, episode two of the ongoing series covering the exciting startups from the AWS ecosystem. We talk about data analytics. I'm your host, John Furrier of theCUBE, and we have Javier de La Torre, who's the founder and chief strategy officer of CARTO, which is doing some amazing innovation around geographic information systems, or GIS. Javier, welcome to theCUBE for this showcase. Thank you. Thank you for having me. So, you know, one of the things that you guys are bringing to the table is spatial analytics. Data that now moves into spatial relations, which is, you know, we know about geofencing, you're seeing more data coming from satellites, ground stations, you name it, things are coming into the market from a data perspective that's across the board. And geo is one of them, GIS systems. This is what you guys are doing, the rise of SQL in particular with spatial. This is a huge new benefit to the world. Can you take a minute to explain what CARTO is doing and what spatial SQL is? Sure. Yeah, so like you said, like data, obviously we know it's growing very fast and it's going to now be leveraged by many organizations in many different ways. But there's one part of data we call one dimension that is location. We like to say that everything happens somewhere. So therefore everything can be analyzed and understood based on the location. So we like to put an example, if all your neighbors get an alarm in their homes, the likelihood that you will get an alarm increases, right? So that's obvious. We are all affected by our surroundings. What is spatial analytics, this type of analytics does is try to uncover those spatial relations so that you can model, you can predict where something is going to happen or you're like, or optimize it, you know, like where else you want it to happen, right? So that's at the core of it. Now, this is something that as an industry has been done for many years, like the GIS or geographic information systems have existed for a long time. But now, and this is what CARTO really brings to the table, we're looking at really democratizing it so that it's in the hands of any analyst. Not our vision is that you don't need to go five years to a geography school to be able to do this type of a spatial analysis. And the way that we want to make that happen is what we call with the rise of a spatial SQL. We add these capabilities around a spatial analytics based on the language that is very, very popular for analysts, which is SQL. So what we do is enables you to do this spatial analysis on top of the well-known and well-used SQL analytics. It's interesting, the cloud native and the cloud scale wave, and now data as code has shown that the old school, the old guard, the old way of doing things. You mentioned data warehousing, okay, as one, BI tools in particular have always been limited. And the scope of the limitation was the environment was different. You have to have domain expertise, rich knowledge of the syntax. Usually it's for an application developer, not for like real time and building it into the CICD pipeline, or just from a workflow standpoint, making it available, the so-called democratization. This is where this connects. And so I got to ask you, what are you most excited about in the innovations at Carto? Can you share some of the things that people might know about or might not know about that's happening at Carto that takes advantage of this cloud native way because companies are now on this bandwagon? Yeah, no, it is. And cloud native analytics is probably the most disruptive kind of trend that we've seen over the few years. In our particular space, on a spacial, it has tremendous effects on the way that we provide our service. So I like to highlight four main reasons why cloud analytics, cloud native is super important to us. So the first one is obviously it's scalability. The working with the sizes of data that we work now in terms of location was just not possible before. So for someone that is performing now analysis on an autonomous car, or you're like, there has any kind of a sensorized GPS on a device and it's collecting billions, hundreds of billions of points. If you want to do analysis on that type of data, cloud native allows you to do that in a scalable way. But it also is very cost-effective. That is very something that you'll see very quickly when your data grows a lot, which is that this computing storage separation, the idea that is to store your data at cloud prices and then use them with these data warehouses that we were describing makes for a very, very cost-effective solution. But then there is other two, obviously one of them being SQL and the spacial SQL that means, we like to say that SQL is becoming the lingua franca for analytics. So it's used by many products that you can connect through the usage of SQL. But I think you're like, you're coming towards, what I think is even more interesting. It's like in the cloud, the concept like we all are serving the, we are all living in the same infrastructure enables us that we can distribute spacial datasets to a customer that they can join it on their database on SQL without having to move the data from one another, like in the case of Redshift on Amazon Redshift, the carto connects and using something called a spectrum, we can connect live to data that is stored on S3. I think that is going to disrupt a lot the way that we think about data distributions and how cost-effective it is. I think it has a lot of your potential on it. And in that sense, what Cartu is providing on top of it in the format of formats like Parquet which is a very popular big data format, we're adding geo-parquet. We are specializing in this big data technology for doing the spacial analysis. And that to me, it is very exciting because it's putting some of the best tools at the hands of doing the spacial analytics for something that we were not able to do before. So to me, this is one area that I'm very, very excited about. And that's, I want to back up for a second. So you mentioned Parquet and the standards around that format. And also you mentioned Redshift. So let me get this right. So you're saying that you can connect into Redshift. So I'm a customer and I have Redshift, I'm using it, I got my S3 I'm using Redshift for analyst. You're saying you can plug right in to Redshift. Yes. And this is a very, very, very important part because what Cartu does is leverage Redshift computing infrastructure to essentially do all the analysis. So what we do is we bring a spacial analysis where the data is where Redshift is versus in the past, what we will do is take the data where the analysis was. And that's it, it's at the core of cloud native. Yeah, this is really where I see the exciting shift where data as code now becomes a reality is that you bring the, it's a redefined architecture. The script is flipped. The architecture has been redefined. You're making the data move to the environments that needs to move when it has to. If it doesn't have to move, you bring compute to it. So you're seeing new kinds of use cases. So I have to ask you on the use cases and examples for Cartu AWS customers with spacial analytics, what are some of the examples on how your clients are using cloud native spacial analytics or Cartu? So one, for example, that we've seen a lot on the AWS ecosystem obviously because of its reach and its position, we work together with another service in the AWS ecosystem called Amazon location. So that actually provides you access to maps and SDKs for navigation. So imagine that you are like a company like that is delivering food or any other goods in the city. You have like hundreds or thousands of drivers around the city moving, doing all these deliveries. And each of these drivers, they have an app and they are collecting actively their location, their position, right? So you get all the data and then it gets stored on something like a Redshift data cluster and there's different architectures in there. But now you essentially have like a full log of the activity that is happening on the ground from your business. So what Cartu does on top of that data is you connect your data into Cartu and now you can do analysis, for example, for finding out where you should maybe place another distribution center for optimizing your delivering routes. Or like if you're on the restaurant business where you might want to add a new dark kitchen, right? So all this type of analysis based on since I know what you're doing in your operations, I can post analyze the data and then provide you a different way that you can think about serving and solving your operation. So that's an example of a great use case that we're seeing right now. So talk to me about the traditional BI tools out there because you mentioned earlier they lack the specific capabilities. You guys bring that to the table. What about the scalability limitations? Can you talk about where that is? Is there limitations there? Obviously, if they don't have the capabilities they can't scale, that's one. But as you start plugging into Redshift scale and performance matters, what's the issue there? Can you unpack that a little bit real quick? Yeah, it goes back to the peculiarities of the spatial data, location data. Like in the use case, like I was describing you very quickly are going to end up with really a lot of your like terabytes if not terabytes of data very quickly if you start integrating all this data because it gets created by sensors. So volumes in our world, that tends to grow a lot. Now, so when you work with BI, with BI tools there's two things that you have to take in consideration. BI tools are great for seeing things. Like for example, if all you want to see is where your customers are, a BI tool is great. Seeing, creating a map and seeing your customers that's totally in the world of BI. But if you want to understand why your customers are there or where else could they be, you're going to need to perform what we call a spatial analysis. You're going to have to create a spatial model. You're going to have to, and for that BI tools will not give you that. That's one side. The other talks about the volumes that I was describing. Most of these BI tools can handle certain aggregations. Like for example, if you are reading if you're connecting your, let's say 10 billion dataset to a BI tool, the BI tool will do some aggregations because you cannot display 10,000 rows on a BI tool. And that's okay to get aggregations and that works. But when it comes to a map, you can not aggregate the data on the map. You actually want to see all the data on the map. And that's what Carto provides you. It allows you to make maps that sees all the data not just aggregated by county or by all other public area. You see all your data on the map. You know, it's interesting is that location based service has been around for a long time. You know, when mobile started even hitting the scene, you saw it get better mashups, Google maps, all this Google API mashups, things like that. You know, developers are used to it but they could never get to the promised land on the big data side because they just didn't have the compute. But now you add in geo fencing, geo information. You now have access to this new edge like data, right? So I have to ask you on the mobile side, are you guys working with any 5G or edge providers? Because I can almost imagine that the spatial equation gets more complicated and more data full. When you start blowing out edge data, like with 5G and you got more things happening at the edge, it's only going to fill in more data points. Can you share how that use case is going with mobile, mobile carriers or 5G? Yeah, it's totally the case. Well, first, even before we are there, we're actually helping a lot of tokens on actually planning the 5G deployment. Where do you place your antennas is a very, very important topic when you're like talking about 5G. Because you know, like 5G networks require a lot of density. There's a lot about like, okay, where do I start deploying my infrastructure to ensure the customers, you know, like, have the best service in the places where I wanna go first. You mean like the RF maps? Like, understanding how RF propagates? Well, that's one signal, but the other is like, imagine that your token is more interested on, you know, like on, let's say, on a certain color consumer profile. Like, you know, like young people, you know, that are using one type of service. Well, we know where these demographics color lives. So you might want to start color deploying your 5G in those areas, right? Versus if you go to more color commercial and more color residential areas, there might be other demographics. So there's one part around market analysis. Then the second part is, once these 5G networks are in place, you're right, I mean, one of the promises that they call like these new technologies give us is because the network is much smarter, you can have all these edge cases. There's much more location that data that can be collected. So what we see now is a rise of, on the amount of what we call telemetry. That for example, the IoT space can make around location. And that's now enabled because of 5G. So I think 5G is going to be one of those trends that are going to make like more and more data coming into, I mean, more location data available for analysis. So how does that, I mean, this is a great conversation because everyone can realize they're at a stadium and they see the multiple bars but they can't get banned with. So you either got to get a backhaul problem or not enough signal. Everyone knows when they're driving their car. They know they can relate to the consumer side of it. So I get how the spatial data grows. What's the impact to CARTA and specifically the cloud because if you have more data coming in, you need the actionable insight. So I can see there the use case. Oh, put the antenna here. That's an actionable business decision. More content, more revenue, more happy customers. But where else is the impact to you guys and the spatial piece of it? Yeah. Well, I mean, like there's many, many factors, right? So one of them, for example, on the, on the technical one of the things where we realize impact is that it gives the visibility to the operator, for example, around the quality of service. Like, okay, it's my, are my customers getting the quality of services where I want? Or like you said, like if there's now suddenly a concert, the quality of service in one particular area is dropping very fast. So the idea of like being able to now in real time kind of like detect location issues. Like I'm having an issue in this place. That means that then now I can act, I can drive a van there, put more capacity, et cetera. So I think the biggest impact that we are seeing and we're going to see on the upcoming years is that, you know, like more and more use cases going towards real time. So where, you know, like before it was like, well, now that it has happened, I'm going to analyze it. I'm going to look at, you know, like how I could do better next time towards a more of like an industry where hard to ourselves, we are embedded in more real time type of, you know, like analytics where it's, okay, if this happens, then do that, right? So it's going to be more personalized at the level that like in a co-collect environment, it has to be part of a full-collect pipeline, kind of like type of analysis that it's already programmatically prepared to act on real time. That's great. And it's good segue to my next question. As more and more companies adopt cloud native analytics, what trends are you seeing out of the key to watch? Obviously you're seeing more developers coming on site on the scene, open source is growing. What's the big cloud native analytics trends for Carto and Geographic information? Yeah. So I think we were talking before the cloud native, now it's unstoppable. But one of the things that we are seeing that still needs to be developed and we are seeing now progress is around a standardization, for example, around like data sets that are provided by different providers. What I mean with that is like, you as an organization, you're going to be responsible for your data lake that you create on your cloud, right? On S3 or in, you know, I can, and then you're going to have a computing engine like Redshift and you're going to have all that set up. But then you also going to have to think about like, okay, how do I ingest data from third party providers that are important for my analysis? So for example, Carto provides a lot of demographics, human mobility, we aggregate them, we clean up and prepare a lot of spatial data so that we can enrich your business. So for us, how we deliver that into your cloud native solution is a very important factor. And we haven't seen yet enough standardization around that. And that's one of the things what we are pushing, you know, with the concept of geo-parquet of you and standardizing that body. That's one. Then there is another, this is more what I'd like to say that, you know, we are helping companies figure out their own geographies. What we mean by that is like most companies, when they start thinking about like, how they interact on the space, on the location, some of them will work like by seed codes and other by cities, they organize their operations based on a geography, you know, or technically what we call a geographic support system. Well, nowadays like the most advanced companies are defining their geographies in a continuous spectrum. What we call global grid system or spatial indexes that allows them to understand the business not just as a set of regions, but as a continuous space. And that is now possible because of the technologies that we are introducing around spatial indexes at the cloud native infrastructure. And it provides a great way to match data resources and operate at the scale. To me, those two trends are gonna be like very, very important on because of the capabilities that cloud native brings to our spatial industry. So it changes the operation. So it's data as ops, data as code is data ops. Like infrastructure as code means cloud DevOps. So I got to ask you, because that's cool. Spatial index is a whole nother way to think of it. Rather than you go hyper local, super local, you get local zones for AWS and regions. Things are getting down to the granularity of levels. I see that. So I have to ask you, what does data as code mean to you? And what does it mean to character? Because you're kind of teasing it this new way because it's redefining the operations, the data operations, data engineering. So data as code is real. What does that mean to you? No, I think we already seen it happening to me and to Cartu, what I will describe data as code is when an organization has moved from doing analysis after the fact, you know, where they're like post like analysis in a way to where they actually kind of put in analytics on their operational cycle. So then they need to really code it. They need to make those, these analysis, put them and insert them into the boot, into the architecture bus, if you want to say of the organization. So if I get a customer and happens to be in this location, I'm going to trigger that. And then this is going to do that. Or if this happens, I'm going to need to open up. And this is where if an organization is going to react in the more real time and we know that organizations need to travel that direction, the only way that they can make that happen is if they operationalize analytics on their daily operations. And that can only happen with data as code. And that's interesting. Look at ML Ops, AI Ops, people talk about that. This is data. So in developers meets operations, that's the cloud, data meets code, that's operations, that's data business. You got it. Add to that the spatial with Cartu and we got it. Because every piece of data now is important and the spatial is key. Real quick, before we close out, what is the index thing? Explain the benefit real quick of a spatial index. Yes, so the spatial index is, well, everybody can understand how we organize societies politically, right? Our countries, you have like states and then you have like counties and you have all these different kind of what we call administrative boundaries, right? That's a way that we organize information too, right? A spatial index is when you divide the world not in administrative boundaries, but you actually make a grid. Imagine that you just actually make a grid of the world, right? And you make that grid so that in every cell, you can then split it into, let's say for example, four more cells. So you now have like an organization, you can just split the world in a grid that you can have multiple resolutions. Think like Google Maps. When you see the entire world, but you can zoom in and you end up seeing like one particular place. So that's one thing. So what a spatial index allows you is to technically put your location not based on a coordinate, but actually on one grid place, on an index. And we use that then later to correlate, let's say your data with someone else data, because we can use what we call is a spatial indexes to do joins very, very fast. And we can do a lot of operations with it. So it is a new way to do computing, spatial computing based on this type of indexes. But for more than anything for an organization, what a spatial index allows is that you don't need to work on zip codes or in boundaries, on artificial boundaries. I mean, your customer doesn't change because it goes from this place to the road to the other side of the road. It's the same place. It's an arbitrary location. It's a spatial index. Break out all of that. You know, like you break with your zip code, you break and you essentially have a continuous geography that actually is a much closer look up to the reality. It's like the forest and the trees and the bark of the tree. You can see everything. Good way of looking at it. Great to have you on and in real quick closing, give a quick plug for the company. Summarize what you do, what you look into, how many people you got when you're hiring, what's the key goals for the company? Yes, sure. So Carto is a company now, we're around 200 people. Our vision is that a spatial analytics is something that every organizations will do. So we really try to enable organizations with the best data and analysis around a spatial and we do all that cloud native on top of your data warehouse. So what we are really enabling these organizations is to take that cloud native approach that they're already embracing it also to a spatial analysis. I'll be a founder, chief strategist, Carto. Great to have you on data as code. All data is real, all data has impact, operational impact with data is the new big trend. Thanks for coming on and sharing the company story and all your key innovations. Thank you. Thanks to you. Okay, this is the startup showcase data as code. Season two, episode two of the ongoing series. Every episode will explore new topics and new exciting companies pioneering this next cloud native wave of innovation. I'm John Furrier, host of theCUBE. Thanks for watching.