 Well, and welcome to this special CUBE Conversation. I'm John Furrier, host of theCUBE here in Palo Alto, California. Featuring StarDog is a great hot startup. We've got a great guest, Rob Harris, Vice President of Solutions Consulting for StarDog here, talking about some of the cloud growth, knowledge graphs, the role of data. Obviously, there's a huge sea change. You're seeing real value coming out of this COVID, companies coming out of the pandemic, new opportunities, new use cases, new expectations, highly accelerated shift happening, and we're here to break it down. Rob, thanks for joining us on theCUBE Conversation. Great to be here. So, Rob, I'm excited to talk to you guys about your company and specifically the value proposition. I've been talking for almost since 2007 around graph databases when Neo4j came out and looking at how data would be part of a real part of the developer mindset early on and this is more of the development. Now it's mainstream. You're seeing value being created in graph structures. Okay, not just relational. This has been very well verified. You guys are in this business. So this is a really hot area, lot of value being created. It's cool and it's relevant. So tell us first, what is StarDog doing? What is the company about? Yeah, so we are an enterprise knowledge graph platform company. We help people be successful in standing up knowledge graphs about the data that they have code inside their company and using public data and tying that all together in order to be able to leverage that connected data and really turn it into knowledge through context and their standard. So how did this all come about? This is from a tech standpoint. What was the motivation around this? Because obviously the unstructured wave hit, you're seeing successes like Databricks, for instance, just absolutely crushing it on their valuation and their relevance. You're seeing the same kind of wave hit almost kind of born back on the Hadoop days with unstructured data. Is that a big part of it? Is it just evolution? What's the big driver here? Yeah, no, I think it's a great question. The driver really is these data sets have increased for so many companies, trying to really bring some understanding to it as they roll it out in their organizations. We've tried to just try to centralize it and that hasn't been sufficient in order to be able to unlock the value of most organization's data. So being able to step beyond just pulling everything together into one place, but really putting that context and meaning around it that the graph can do. So that's where we've really got started at back in the day is we really looked at the inference and reasoning part of a knowledge graph. How do we bring more context to understanding that doesn't naturally exist within the data? And that really is how we launched off the product from the very beginning. I got to ask you around the use cases because one of the things that's really relevant right now is seeing a lot of front end development around agile application. DevOps has brought infrastructure as code. You're seeing kind of this huge tsunami of new kinds of applications. One of the things that people are talking about in some of the developer circles and it's kind of hits the enterprises this notion of state because you can have an application calling data but if the data is not addressable and then keeping state in real time and all these kinds of new technical problems. How do you guys look at that when you look at trying to create knowledge graphs because maintaining that level of connection you need data, a ton of it it's got to be exposed and addressable and then dealt with in real time. How do you guys look at that? Yeah, that's a great question. What we've done to try to kind of move the ball forward on this is move past trying to centralize that data into a knowledge graph that is separate from the rest of your data assets but really build a data virtualization layer which we have integrated into our product to look at the data where it is in the applications and the unstructured documents and the structure repositories so that we can observe as state changes in that data and answer questions that are relevant at the time and we don't have to worry about some sort of synchronous process loading information into the graph. So that ability to add that virtualization layer to the graph really enables you to get more of a real time look at your data as it evolves with the organization. Yeah, I definitely want to double click on that and say, but I want to just drop step back and kind of set the table for the folks that aren't getting in the weeds yet on this. There's kind of a specific definition of enterprise knowledge graph. Do you like just quickly define that? What is the enterprise knowledge graph? Sure, yeah, we really see an enterprise knowledge graph as a connected set of data with context. So it's not just storing it like a graph but connecting it and putting meaning around that data through structure, through definitions, et cetera across the entire enterprise. So looking at not just data within a single application or within a single silo but broadly through your enterprise what does your data mean? How is it connected? And what does it look like within context in each other? How should companies reuse their data? Boy, that's a broad question, right? You know, I mean, one of the things that I think is very important is so many companies have just collected data assets over the years, they collect more and more and more. We have customers that have eight petabytes of data within their data lake and they're trying to figure out how to leverage it like actually connecting and putting that context around the data. You can get a lot more meaning out of that old data or the stale data or the unknown data that the people are getting right today. So the ability to reuse the data assets within context and meaning is where we've seen people really be able to make huge lips for in their organization like drug companies be able to get drugs to market faster by looking at older studies they've done where maybe the meaning was hidden because it was an old system and nobody knew what the particular codes and meaning were in context of today. So being able to reuse and bring that forward brings real life application to people solving business problems today. Rob, I got to get your thoughts on something that we always riff on here on theCUBE which is, do you take down the data silos or do you leverage them? And you know, this came up a lot many years ago when we first started discussing containers for instance and then we saw that you didn't have to kill the old to bring in the new. There's one mindset of break silos down, go horizontal scalability on the data, create a data plane, control plane, others saying, hey, you know what, just put a wrapper around those silos. I'm over simplifying, but you get the idea. So how should someone who's really struggling with or not struggling with putting together an architecture around their future plans around dealing with data and data silos specifically because certainly as new data comes in there's mechanisms for that, but as you have existing data silos, what do companies do? What's the strategy in your opinion? Yeah, you know, it is a really interesting question. I was in data warehousing for a long, long time and a big proponent of moving everything to one place. And then I really moved into looking at the data virtualization and realized that neither of those solutions are complete, that there are some things that have to be centralized and moved. The old systems aren't sufficient in order to be able to answer questions or process them, but there are many data silos that we've created with organizations that can be reused. You can leverage the compute, you can leverage the storage that already exists within those. And that's the approach we've taken at Stardoc. We really want to be able to allow you to centralize the data that makes sense, right? To get it out of those old systems that should be shut down from just a monetary perspective, but the systems that have actual meaning or that it's too expensive in order to remove them leverage those data silos. And by letting you have both approaches in the same platform, we hope to make this not an either or architectural decision, which is always the difficult question. Okay, so you got me on that one. So let me just say that I want to leverage my data silos. What do I do? Take me through the playbook. What, if I got the data silos, what is the Stardoc recommendation for me? Sure. So what we generally recommend is you start off with building kind of a model in the lingo we sometimes say ontology or some sort of semantic understanding that puts context around what is my data and what does it mean? And then we allow you to map those data silos. We have a series of connectors in our product that whether it's an application and you're connecting through a REST connector or whether it's a database and you're connecting through ODBC or JDBC map that data into the platform. And then when you issue queries to the Stardoc platform we federate those queries out to the downstream systems and answer as if that data existed on the graph. So that way we're leveraging the silos where they are without you having to move the data physically into the platform. So you guys essentially building a data fabric? We are, yeah, data fabric is really the new term that's been popping up more and more with our customers when they come to us to say how can we kind of get past the traditional ways of doing data integration and unify data in a single place? Yeah. Like we said, we don't think the answer is purely all about moving it all to one big lake. We don't think the answer is all about just creating this virtualization plane but really being able to leverage the best of both. All right, so if you believe that then let's just go to the next level then. So if you believe that they don't have to move things around and to have one specific thing, how does a customer deal with their challenge of hybrid cloud and soon to be multi-cloud? Because that's certainly on the horizon. People want choice. There's going to be architectural. I mean, certainly a cloud operations will be in play but this on-premises and this cloud and then soon to be multiple cloud. How do you guys deal with that question? Yeah, that's a great question. And this is really an area that we're very excited about and we've been investing very heavily in. Is how to have multiple instances of StarDog running in different clouds or on-prem on the cloud coordinate to answer questions to minimize data movement between the platforms. So we have the ability to run either an agent on-prem. For example, if you're running the platform in the cloud or vice versa, you can run it in the cloud. You already have two full instances of StarDog where they will actually co-plan queries to understand where does the data live? Where is it resident? How do I minimize moving data around in order to answer the question? So we really are trying to create that unified data fabric across on-prem or multiple cloud providers so that any of the nodes in the platform can answer a question from any of the data sources. You know, complexity is always the issue. People cost go up when you have complexity. You guys are trying to tame it. This is a huge conversation you bring up multi-cloud and I mean hybrid cloud and multi-cloud. When you think about the IoT edge and you don't want to move data around this is what everyone's saying. Why move it? Why move data? It's expensive to move data, process it where it is and you're going to have this kind of flexibility. So this idea of unification is a huge concept. Is that enough? I mean, and how should customers think about the unification because if you can get there, it almost, it is the kind of the holy grail what you're talking about here. So this is kind of the prospect of having kind of ideal architecture of unification. So take me through that one step deeper. Well, it is kind of interesting because as you really think about unifying your data and really bringing it together, of course it is the holy grail. That's what people have been talking about. Gosh, since I started in the industry over 20 years ago how do I get this single playing view of my data regardless of whether it's physically located or somehow stitched together. But one of the things that, you know our founders really strongly believed on when the start of the company was it isn't enough, it isn't sufficient. There is more value in your data that you don't even know. And unlocking that through either machine learning which is of course we all know it's very hot right now to look at how do I derive new insights out of the data that I already have. Or even through logical reasoning, right? And inference, looking at what do I understand about how that data is put together and how it's created in order to create more connections within the data and answer more questions. All those are ways to grow beyond just unifying your data but actually getting more insights out of it. And I think that is the real holy grail that people are looking for not just bringing all the data together but actually being able to get business value and insights out of that data. Yeah, I'm looking forward. You guys have obviously a pretty strong roster of clients that represent that. But I got to ask you since you brought up the founders of the company obviously having a founders DNA mindset tends to change the culture or drive the culture of the kind of change when they drive the culture of the company. What is the founders culture inside StarDog? What's the vibe there? If you could, what do they talk about the most when they get in that mode of being founders like, hey, you know, this is the North Star. What is the rap like? What's the vibe? Take us through some StarDog culture. Sure. So our three founders came out of the University of Maryland all in a PhD program around semantic reasoning and logical understanding. And being able to understand data and be able to communicate that as easily as possible is really the core and the fiber of their being. And that's what we see continually under discussion every single day. How can we push the limits to take this technology and you get easier to use, more available, bring more insights to the customers beyond what we've seen in the past. And I find that really exciting to be able to constantly have conversations about how do we push the envelope? How do we look beyond even what Gartner says is five or eight years in the future but looking even further ahead. So they're into this whole data scene then big time. They are, that they are very active in the conferences and posts and all that great stuff. Agility, they love this agility. They got to love DevOps. I mean, if you're into this knowledge graph scene. So I got to ask you, what's the machine learning angle here? Obviously AI, we know what AI is. AI is essentially a combination of many things, machine learning and other computer science and data access. What is the secret sauce behind the machine learning and the vibe and the product of StarDog? Yeah, a lot of times the way that we leverage machine learning or the way that we look at it is, how do we create those connections between data? So you have multiple different systems and you're trying to bring all that data together. And it's not always easy to tell, is this raw Paris the same as that raw Paris? Is this product the same as that product? So when it's possible, we will leverage keys or we'll leverage very systematic type of understanding of these things are the same, but sometimes you need to reach beyond that. And that's where we leverage a lot of machine learning within the platform, looking at things like linear regression or other approaches around the graph, you know, connectivity analysis, page rank, things like that to say, where are things the same so that we can build out that connections and that connectivity as automatically as possible? You know, I give a lot of talks on theCUBE, also now the new, new clubhouse app where people are talking about misinformation. Obviously we're in the media business. We love the digital network effect, everything's networked, it's a network economy. We're starting to see this power of information and value, you guys call it the knowledge graph. So I got to ask you, when you look at this kind of future where you have this complexity in the network effect, how are you guys looking at that data access? Because if you don't have the data, you're not going to have that insight, right? So you need to have that network connection. Is that a limitation for companies? Is that an, because usually people aren't necessarily, their blind spot is their data, or their lack of their data. So having things networked together is going to be more the norm in the future. How do you guys see that playing out? Yeah, I think you're exactly right. And I think that as you look beyond where we are today, a lot of times we focus today on the data that a company already has. What do I know, right? What do I know about you? What, how do I interact with you? How have I interacted with you? I think that as we look at the future, we're going to talk more about data sharing, about leveraging publicly available information, about being able to take these insights and leverage them not just within the walls of my own organization, but being able to share them and work together with other organizations to bring a better understanding of you as a person or as a consumer that we could all interact with. Yeah, you're absolutely right. Metcots law still holds true that more network connections bring more value. I certainly see that growing in the future probably more around more data sharing and more openness about leveraging publicly available. You know, it's interesting, you mentioned you came from a data warehouse background. I remember when I broke in the business mill 30 years ago when I started getting computer science, you know, it was, it was, it was different having a product and an enabling platform. You guys seem to have this enabling platform where there's no one use case. I mean, you have an unlimited use case landscape. You could do anything with what you guys have. It's not so much, I mean, there's low hanging fruit. So I got to ask you if you have that enabling platform, you're creating value for customers. What are some of the areas you see developing like now in terms of low hanging fruit and where's the possibilities? How do you guys see that? I'm sure you've probably got a tsunami of activity around corner cases from media to every vertical. We do, and that's, you know, part of the exciting part of this job part of the exciting part of Knowledge Graphs in general is to see all the different ways that they are allowed to use. But when we do see some use cases repeated over and over again, risk management is a very common one. How do I look at all the people and the assets with an organization, the interactions they have to look at hotspots for risk that I need to correct within my organization? For the pre-commercial pharma that has been a very, very hot area for us recently. How do we look at all the research that's available with an organization that's publicly available in order to accelerate drug development in this post COVID world that's become more and more relevant for organizations to be able to move forward faster in the kind of the bio industry and my sciences. That's a use case that we've seen repeated over and over again. And then this growing idea of the data fabric, the data fabric looking at metadata within the organization to improve data integration processes to really reduce the need for moving data around the organization as much. Those are use cases we've seen repeated over and over again over the last year. Awesome, Rob, my last question before we wrap up is for the solution architect that's out there that has got a real tall order. They have to put together a scalable organization, people, process and technology around a data architecture that's going to be part of the next gen, the next gen next level activity. They need headroom for IoT edge, industrial edge and all use cases. What's your advice to them as they have to look out and start thinking about architecture in their organization? Yeah, that's a great question. I really think that it's important to keep your options open as the technology in the space continues to evolve. It's easy to get locked into a single vendor or a single mindset. I've been an architect most of my career and that's usually a lot of the pitfalls. Things like a knowledge graph are open and flexible. They adhere to standards which then means you're not locked into a single vendor and you're allowed to leverage this type of technology to grow beyond what you originally envisioned. Thinking about how you can take advantage of these modern techniques to look at things and not just keep repeating what you've done in the past, the sins of the past have a, a lot of times do reappear. So fighting against that as much as possible is my encouragement. Awesome, great insight and I love this area. I know you guys got a great trend you're riding on. Very cool, very relevant. Final minute, just take a quick minute to give a plug for the company. What's the business model? How do I deploy this? How do I get the software? How do you charge for it? If I'm going to buy this solution or engage with Stardow, what do I do? Take me through that. Sure. Yeah, we are like a, you've said through this whole thing we are an enterprise knowledge graph platform company. So we really help you get started with your business leveraging and using a knowledge graph within your organization. We have the ability to deploy on-prem. We have the ability to deploy on the cloud. We're in the AWS marketplace today. So you can take a look at our software today. We generally are subscription based based on the size of the install and we are happy to talk to you anytime. Just drop by our website and reach out. We'll get to talk to you soon. Rob, great. Thanks for coming on. I really appreciate it. Gradient said, looking forward to seeing you in person when we get back to real life. So hopefully the vaccines are coming on. Thanks to companies like you guys providing awesome analytics and intelligence for these drug companies and pharma companies. Now you have a few of them on your client roster. So congratulations. Looking forward to following up. Great, great area. Cool and relevant data. Architecture is changing. Some of it's broken. Some of it's being fixed. StarDog is one of the hot startups scaling up beautifully in this new era of cloud computing meets applications and data. So I'm John Furrier at theCUBE. This is a CUBE conversation from Palo Alto, California. Thanks for watching.