 Welcome everyone to theCUBE's presentation of the AWS startup showcase, Data as Code. This is season two, episode two of our ongoing series covering exciting startups in the AWS ecosystem to talk about data and analytics. I'm your host, Lisa Martin. I have a CUBE alumni here with me, Sokka Tsarav, the CEO and founder of Nexla. He's here to talk about a future of automated data engineering. Sokka, welcome back, great to see you. Lisa, thank you for having me. Pleasure to be here again. Let's dig into Nexla's mission, ready to use data in the hands of every user. What does that mean? That means that, you know, every organization, what are they trying to do with data? They want to make use of data. They want to make decisions from data. They want to make data a part of their business, right? The challenge is that every function in an organization today needs to leverage data, whether it is finance, whether it is HR, whether it is marketing sales or product. The problem for companies is that for each of these users, each of these teams, the data is not ready for them to use as it is. There is a lot that goes on before the data can be in their hands. And it's in the tools that they like to work with. And that's where a lot of data engineering happens today. I would say that is by far one of the biggest bottlenecks today for companies in accelerating their business and being truly data driven. So talk to me about what makes Nexla unique. When you're in customer conversations, as every company these days in every industry has to be a data company, what do you tell them about what differentiates you? Yeah, one of the biggest challenges out there is that the variety of data that companies work with is growing tremendously. You know, every SaaS application you use becomes a data source. Every type of database, every type of user event, anything can be a source of data. Now, it is a tremendous engineering challenge for companies to make the data usable. And the biggest challenge there is people. Companies just cannot have enough people to write that code to make the data engineering happen. And where we come in with a very unique value is how to start thinking about making this whole process much faster, much more automated. At the end of the day, Lisa, time to value and time to results is by far the number one thing on top of mind for customers. Time to value is critical. We're all thin on patience these days, whether we're in our consumer lives or our business lives. So being able to get access to data to make intelligent decisions, whether it's on something that you're going to buy or a product or service you're going to deliver is really critical. Give me a snapshot of some of the users of Nexla. Yeah, the users of Nexla are actually across different industries. One of the many, one of the interesting things is that the data challenges, whether you are in financial services, whether you are in retail e-commerce, whether you are in healthcare, they are very similar, is basically getting connected to all these data systems and having the data. Now, what people do with the data is very specific to their industry. So for example, within the e-commerce world or retail world itself, companies from the likes of Bed Bath Beyond and Forever 21 and Poshmark, which are retailers or e-commerce companies, they use Nexla today to bring a lot of data in. So do delivery companies like Do dash and Instacart. And so do, for example, logistics providers like Narva or customer loyalty and customer data companies like Yachtwo. So across the board, for example, just in retail we cover a whole bunch of companies. Got it. Now let's dig into, let's talk about the future of automated data engineering. Talk to me about data engineering. What is it? Let's define it and crack it open. Yeah, data engineering is, I would say, by far one of the hottest areas of work today, the one of the hardest people to hire if you're looking for one. Data engineering is basically all the code, the process and the people that is basically connecting to their system. So just to give a very practical example, right? For somebody in e-commerce, let's say a take off case of Do dash, right? It's extremely important for them to have data as to which stores have what products, what is available. Is this something they can list for people to go and buy? Is this something that they can therefore deliver, right? This is data that changes all the time. Now imagine them getting data from hundreds of different merchants across the board. So it is the task of data engineering to then consume that data from all these different places, different formats, different APIs, different systems, and then somehow unify all the data so that it can be used by the applications that they are building. So data engineering in this case becomes taking data from different places and making it useful. Again, back to what I was talking about, ready to use data. It is a lot of code. It's a lot of people, not just that. It is something that runs every single day. So it means it has monitoring. It has reliability. It has performance. It has every aspect of engineering, as we know, going into it. You mentioned it's a hot topic, which it is, but it's also really challenging to accomplish. How does Nexla help enable that? Yeah, data engineering is quite interesting in that one, it is difficult to implement the necessary set of pieces, but it is also very repetitive at some level, right? I mean, when you connect to, say, 10 systems and get data from them, that's not the end of it. You have 10 more and 10 more and 10 more. And then at some point, you have thousands of such data connectivity and data flows happening, it's hard to maintain them as well. So the way Nexla gets into the whole picture is looking at what can we understand about data? What can we observe about the data systems? What can we learn from that? And then start to automate certain pieces of data engineering. So that we are helping those teams just accelerate a lot faster. And I would say comes down to more people being able to do these tasks, rather than only very, very specialized people. More people being able to do the tasks, more users, kind of democratization of data really there. Can you talk to us in more detail about how Nexla is automating data engineering? Yeah, I think this is best shared through a visual. So let me walk you through that a little bit as to how we automate data engineering, right? So if you think about data engineering, three of the most core components are many parts to it, but three of the most core components of that are integrating with data systems, preparing and transforming data and then monitoring that, right? So automating data engineering happens in three different ways. First of all, connecting. Connecting to data is basically about the gateway to data, the ability to read and write data from different systems. This is where the data journey starts, but it is extremely complex because people have to write code to connect to different systems. One part that we have automated is generating these connectors so that you don't have to write code for that. Also making them bi-directional is extremely valuable because now you can read and write from any system. The second part is that the gateway, the connector has read the data, but how do you represent it to the user so anybody can understand it? And that's where the concept of data product comes in. So we also look at auto-generating data products. These become the common language and entity that people can understand and work with. And then the third part is taking all this automation and bringing the human in the loop. No automation is perfect. And therefore bringing the human in the loop means that somebody who is an expert in data who can look at it and understand it can now do things which only data systems experts were able to do before. So bringing in that user of data directly into the picture is one important part, but let's not forget data challenges are very diverse and very complex. So the same system also becomes accessible to the engineers who are experts in that. And now both of these can work together while an engineer will come through API's and SDK and Kamala interfaces, a data user comes in through a nice no-code user interface. And all of these things coming together are what is accelerating back to that time to value that really everybody cares about. So if I'm in marketing and I'm a data user I'm able to have a collaborative workflow with the data engineer. Yeah, yeah. For the first time that is actually possible. And everybody's focuses on their expertise and their know-how. So somebody who for example in financial services really understands portfolio and transactions and different type of asset classes they have the data in front of them. The engineers who understand the underlying real-time data feeds and those they are still involved in the loop but now they are not doing that back and forth. As the user of data I'm not going to the engineer and saying, hey, can you do this for me? Can you get the data here? And that back and forth is not only time taking is frustrating and the number one hold back. Right, yeah. And that's time that nobody has to waste as we know for many reasons. Talk to me about when you look into your crystal ball which I'm sure you have one. What is the future of data engineering from Nexla's perspective? You talked about the automation. What's the future hold? I think the future of data engineering becomes that we up level this at a point where companies don't have to be slowed down for it. I think a lot of tooling is already happening. The way to think about this is that here in 2022 if we think that our data challenges are like X they will be a thousand X in five years. I mean, this complexity is just increasing very rapidly. So we think that this becomes one of those fundamental layers. And as I was saying maybe the last time this is like the road. You don't feel it. You just move on it. You do your job. You build your products. You deliver your services as a company. This just works for you. And that's where I think the future is. And that's where I think the future should be. We all need to work towards that. We're not there yet. Not there yet. A lot of potential, a lot of opportunity and a lot of momentum. Speaking of momentum, I want to talk about data mesh. That is a topic of a lot of excitement, a lot of discussion. Let's unpack that. Yeah, I think the idea that data should be democratized that people should get access to the data. And it's all coming back to that sort of basic concept of scale. Companies can scale only when more people can do the relevant jobs without depending on each other, right? So the idea of data democratization has been there for a long time. But recently in the last couple of years, the concept of data mesh was introduced by Zamaq Ligani at ThoughtWorks. And that has really caught the attention of people and the imagination of leadership as well. The idea that data should be available as a product, that democratization can happen. What is the entity of the democratization that's data presented as a product that people can use and collaborate is extremely powerful. I think a lot of companies are gravitating towards that. And that's why it's exciting. This is promising a future that is possible. So Salkit, speaking of data products, we talked a little bit about this last time, but can you really help us understand, see, smell, touch, feel what a data product is and give us that context? Yeah, absolutely. I think best to orient ourselves with the general thinking of how we consider something as a product, right? A product is something that we find ready to use. For example, this table that I'm using right now, made out of raw materials, wood, metal, screws, somebody designed it, somebody produced it, and I'm using it right now. When we think about data products, we think about data as a raw material. So for example, a spreadsheet, an EPI, a database query, those are the raw materials. What is a data product is something that further enriches and enhances that entity to be much more usable, ready to use, right? Let me illustrate that with a little bit of a visual, actually, and that might help, okay? The idea of the data product, and this is how a data product looks like in NextLap for a user to write as you see. The concept of a data product is something that first of all, it's a logical entity. This simply means that it's not a new copy of data, just like containers or logical compute units. These data products are logical entities, but they represent data in the same consistent fashion regardless of where the data comes from, what format it is in. They provide the user the idea of what the structure of data is, what the sample data looks like, what the characteristics of data are. It allows people to have some documentation around it. What does the data mean? What do these attributes mean? And how to interpret them? How to validate that data? Something that users often know in an industry. How is my data looking like? Well, this value can never be negative because it's a price, for example, right? Then the ability to take these data products that we automate by generating, as I was mentioning earlier, automatically creating these data products, taking these data products to create new data products. Now, that's something that's very unique about data. You could take data about an order from a company and say, well, the order data has an order ID and a user ID, but I need to look up shipping address. So I can combine user and order data to get that information in one place. So creating new data products, giving people access. Hey, I've designed a data product. I think you'll find it useful. You can go use that as it is. You don't have to go from scratch. So all of those things together make a data product, something that people can find ready to use again. And this is also usable by the, again, that example where I'm in marketing or I'm in sales. This is available to me as a general user. As a general user in the tool of your choice. So you can say, oh, no, I'm most familiar with using data in a spreadsheet. I would like it there. Or I prefer my data in a tab lower or a looker to visualize it and you can have it there. So these data products give multiple interfaces for the end user to make use of it. Got it. I like it. You're meeting the user where they are with relevant data that helps them understand so much more contextually. I'm curious, when you're in customer conversations, customers that come to you saying, Saka, we need to build the data mesh. How is Nexla relevant? What is your conversation like? Yeah. When people want to build a data mesh, they're really looking for how their organization will scale into the future. There are multiple components to building a data mesh. It's a tooling part of it, the technology portion. There are people and processes, right? I mean, unless you train people in certain processes and say, hey, when you build a data product, make sure you have taken care of privacy or compliance to certain rules or who do you give access to is something you have to follow some rules about. So we provide the technology component of it. And then the people and processes something that companies, then as they adopt and do that, right? So the concept of data product becomes core to building the data mesh, having governance on it, having all this be self-service and essential part of that. So that's where we come into the picture as a technology component to the whole story. And working to deliver on that mission to getting data in the hands of every user. You mentioned, I want to dig into in the last few minutes here that we have the target audience. You mentioned a few by name, big names customers that Nexla has. I had not heard retail, I heard e-commerce, I think I heard logistics. But talk to me about the target customer for Nexla. Any verticals in particular or any company sizes in particular as well? Yeah, one of the top three banks in the country is a big user of Nexla as part of their data stack. We actually sit as part of their enterprise-wide AI platform providing data to the data scientists. We're not allowed to share their name, unfortunately, but there are multiple other companies in asset management area. For example, they work with a lot of data in markets, portfolio, and so on. The leading medical devices companies using Nexla. Data scientists there are using data coming in real time or streaming from medical devices to train and combine that with other data to do sort of clinical trial related research that they do. We have the companies, for example, LinkedIn is a Nexla customer. LinkedIn is by far the largest social network. Their marketing team leverages Nexla to bring data from different types of systems together as well. So are companies in education space like Nerdy is a public company that uses Nexla for student enrollment education data as they collaborate with school districts, for example. There are companies across the board in marketing, live ramp, for example, uses Nexla. So we are from who uses Nexla is today mostly meant to large to very large enterprises today, leverage Nexla as a very critical component and often mission critical data for which they leverage us. Do you see that changing anytime soon as every company these days has to be a data company? We expect that as consumers, whether it's my grocery store or my local coffee shop that they've got to use data to deliver me that personalized experience. Do you see the target audience kind of shifting down to more into mid-market SMB space for Nexla? Oh yeah, absolutely. Look, we started the journey of the company with the thinking that the most complex data challenges exist in the large enterprise. And if we can make it no core self-serve easy to use for them, we can bring the same high-end technology to everybody. And this is exactly why we recently launched in the Amazon marketplace. So anybody can go there, get access to Nexla and start to use it. And you will see more and more of that happen where we will be bringing even some free versions of our product available. So you're absolutely right. Every company needs to leverage data. And I think people are getting much better at it. Especially in the last couple of years, I've seen that teams have become much more sophisticated. Yes, even if you're at a coffee shop and you're running campaigns, getting people, Yelp reviews and so on, all this data that you can use and understand better your demographic, your customer and run your business better. So as one day, yes, we will absolutely be in the hands of every single person there. A lot more opportunity to delight a lot more consumers and customers. Socket, thank you so much for joining me on the program during the startup showcase. You did a great job of helping us understand the future of automated data engineering. We appreciate your insights. Thank you so much, Lisa. It's a pleasure talking to you. Likewise, for Socket Sareb, I'm Lisa Martin. You're watching theCUBE's coverage of the AWS startup showcase season two, episode two. Stick around, more great content coming up from theCUBE, the leader in hybrid tech event coverage.