Dave Mariani: Hi, everyone. Welcome to another episode of AtScale’s Data Driven podcast. And today’s special guest is Justin Borgman. Justin is a chairman, CEO and founder of Starburst. Justin, welcome to the podcast. Thanks,
Justin Borgman: Dave. Great to be here.
Dave Mariani: Yeah. You know, Justin, so listeners, we, Justin and I have known each other for a long time as fellow entrepreneurs in this new sort of big data space. So, there’s lots to talk about. So we’re gonna have a fun time reminiscing a little bit about just how we all got started, and, and then, and then dive more deeply into what’s happening in the realm of data. Lots of juicy topics here, but Justin, why don’t you just tell the listeners a little bit about yourself, about Starburst, your path into analytics. So, how did you get to where you are today? Let’s just start there.
Justin Borgman: Sure. Yeah. So, obviously Justin Borgman, co-founder and CEO of Starburst. For me, the journey in big data and analytics goes all the way back to early 2010. And back then, I had, I was actually in grad school at the time and had just met a young professor at Yale. His name was Daniel Abadi. And Daniel, his PhD dissertation actually earlier became Vertica. So he had already had a, you know, a big impact on the field, you know, at, at a very young age and had just become a professor at Yale. And one of the first papers that he and his researchers wrote was something called Hadoop db. And the idea here was turning Hadoop into really a data warehouse. And, I thought that was really interesting. I had come from a software engineering background.
Justin Borgman: I was getting a, an MBA at the time, and I said, wow, you know what I, I think this should be a, a business. In fact, I think this idea of doing SQL analytics in a data lake, could be a big thing. and so, so we started our first company together, which became a company called Head Adapt. I actually dropped outta school, to, to do it. And, you know, raised venture and, and built that business over four years, ultimately selling it to Teradata. In 2014, at Teradata, I became a, a VP and GM, responsible for a portfolio of products that was really aimed as sort of figuring out the future of data warehousing analytics. So it was sort of like, emerging technologies, if you will. and so Hadoop was part of that, or at least Teradata’s approach to Hadoop was part of that.
Justin Borgman: But it also gave me a lot of room to explore and really think about, you know, ways that we could, you know, really help Teradata solve their own innovators dilemma. You know, that that classic Clayton Christensen, you know, book Teradata, the poster child for a lot of that. And so, you know, I, I tried to think about how we could disrupt ourselves, and that’s really what led me to a different open source project created at Facebook that was, at the time known as Presto. Today, it’s more commonly known as Trino, but the, the idea was, you know, this is a SQL engine, something that I already had a lot of experience with, but a SQL engine for everything. And that was really interesting to me that you could query data in a variety of different data sources and really turn the data warehousing analytics model inside out.
Justin Borgman: You know, the, the model had always classically been about centralizing everything in one place, and this technology really allowed you to turn that inside out and allow you to query the data where it lives. So I started collaborating with the creators at Facebook. even while I was at Teradata, we started to advance the, the project, add more enterprise capabilities to it. And my thought was, you know, maybe, maybe this could be, you know, a big business for Teradata itself and really, you know, sort of change the direction of the company. I wasn’t successful necessarily at changing the direction of Teradata, but ultimately in 2017, left Teradata and formed Starburst as the company behind this project, and the creators from Facebook, left, Martine Daye and David, and joined me. And, and, you know, off we went. And so that was kind of the, the beginning. And, and of course now we’ve raised over 400 million in Metro Capital. We’re 600 people. a lot has changed in in the last few years, but that’s a little bit about yeah. Kind of how we got started and, you know, have spent, for better or worse, you know, almost 15 years now in, in this space.
Dave Mariani: Yeah. Let’s like, let, let’s just take that apart because, you know, I always get questions from people about, about the sort of the entrepreneurial journey. Yeah. And so our, and our, our backgrounds look very similar. so, in terms of like how we sort of did it, it’s like I started my first company, called Mindshare, you know, yours was adapt. Yep. and, you know, ended up, you know, not making a big success out of it, and selling it for a, a decent amount, but, but selling it to another company, and in that other company was, was really where, you know, was a much bigger company, where we were able to, was able to to do more. but, you know, talk a little bit about just like, you know, you, you raised money twice on an idea mm-hmm. , starting with Adapt, you were still in school, right and, and so you weren’t a proven entrepreneur. I mean, cuz you were still, and you quit school. Like, you’re, like, you’re like Zuckerberg, you know, and and, and Bill Gates. So you followed their footsteps. So, so what was it like to actually get that venture adapt off the ground when you were still in school How’d you do that
Justin Borgman: Yeah, and I will say, you know, this, this is probably not what, you know, parents want to hear if anybody’s listening, but for some reason, when you drop outta school, it actually adds street cred to you. So, you know, it like, it like helps in, in the fundraising process that, that I had dropped out. but yeah, I mean, you’re, you’re right. So first of all, as a, as a first time entrepreneur, I think it is very hard, and I have tremendous respect for anybody who is going down that path trying to get something off the ground, because I think the, the, the secret that most people don’t know is the vast majority fail before they even really get started. You know, unfortunately, mm-hmm. , this, this is a, this is a game where, you know, the odds are against you.
Justin Borgman: And I think, you know, because of that, to me, one of the most important predictors of success is literally the, the will, the grit, the determination of the, of the individual entrepreneur. Like, you have to be willing to have people say no a hundred times to your face and keep going. And that persistence, you know, can eventually pay off. So that was part of our story. I, I didn’t probably know any better that that also helps, I think when you’re young. and, you know, just felt like this was the way the market was going to go. You know, Hadoop was just starting to really gain momentum outside of Yahoo, which I know you obviously spent some time, at, at Yahoo and, and you were now starting to see other companies, deploy it. Cloudera had been funded and it was, you know, creating more, popularity around the project.
Justin Borgman: Hortonworks, you know, came shortly after that. So there was momentum in the market, I think, around Hadoop broadly. And, you know, we had a different approach. We were really thinking about it as a data warehouse, not just scalable storage or scalable batch processing, but like, could this actually be, you know, a sequel data warehouse essentially. And, you know, through persistence and, you know, probably some good luck along the way, we were able to, you know, get some investors to believe in that. But initially it started with angel investors. So, I mean, our first round was finding, angel investors actually local to the Connecticut, you know, new Haven area where we were, we were getting this thing started. Not necessarily a database hub, by the way, . but that’s, that’s, we were getting this off the ground. And, you know, fortunately had, you know, a couple dozen people, individuals, you know, believe in us, give us our first 1.5 million to go start to really move this thing forward. And then, you know, from there, it was Bessemer and Norwest that, that led our series A and then, shortly thereafter actually brought Chris Lynch on the board who, who you obviously know at, at, AtScale. And, he was a tremendous mentor to me as well. I think that’s another piece of advice that I give entrepreneurs is like, find good mentors who can, you know, teach you and show you the ropes. And, and I was fortunate to have a few, but none bigger than than Chris at that point.
Dave Mariani: Yeah. So, listeners, Chris Lynch is the CEO of AtScale. so, so that, it’s definitely, the data community is definitely pretty incestuous and connected. And another, another example of that is, and I didn’t know you from Adam, Justin back when I was at Yahoo, and this was back in 2010. and and I don’t even know how we got acquainted, but somehow we ended up, you know, drinking beers and having dinner in Palo Alto, ta when you were, when you were, forming, adapt at the time.
Justin Borgman: Y yes. And funny enough, I remember in that conversation you were talking about Ola cubes and like you were trying to convince me to go in a slightly different direction, I think, at the time. And, you know, lo and behold that basically became your company, you know, shortly thereafter, which is really cool. So it is funny how all these things intersect and you do find yourself, running into a lot of the same people in this industry over and over,
Dave Mariani: You know But you said something that, that I really, that I really, really rings true, is that you have to have the conviction because, you know, for me personally, it’s trying to fund AtScale, and get it off the ground, get that first, you know, that first, and in my case, it was seed, a seed round. it, you know, it, it probably took almost a year to actually do that. And, and I’ve gotten, I got told no by just about everybody in the valley, except for, a Ryan Floyd of Storm Ventures who had the, who had the foresight to see that yeah. That this was, you know, that this was a new paradigm and running analytics, AtScake, you know, on a new sort of cloud-based infrastructure or a new sort of distributed infrastructure. Cuz cuz we started with the DUP as well.
Dave Mariani: So we were of like minds there. you know, was was gonna be, was gonna be something, where the market would move. but at the same time, right. so, I, I got into Yahoo, through an acquisition, just like you got into Teradata through an acquisition. And so where our stories are also sort of, overlapped not just on a time basis, but also just, for me at least personally being part of Yahoo, which is my first big company where I was a part of, had challenges and had challenges because, you know, it’s, there’s, there’s a lot of people with jobs that I call them the talking heads. and very difficult to sort of break through the talking heads. and, I was also surprised that, you know, that, that at that time, Yahoo didn’t really have a plan for what they wanted to do with the gap, that acquisition.
Dave Mariani: and so you can go two different directions. You can say, well, I’m just gonna just, just sit there and, and rest invest, like a lot of people do. or you can do what you did and what I decided to do, which is like, Hey, look, I got all these resources, I got this, all this big data, I got all these, I got all these customers, why don’t when I go and try to do something crazy and see where it goes And it sounds like you had this similar experience at Teradata when you sort of like started to really, peek behind the covers of Presto, and to see what you could do there.
Justin Borgman: Y totally, and I, I actually think I, I, I totally resonate with what you just said in, in a way, and yes, you can take those two paths and certainly, you know, I know plenty of entrepreneurs that are like, you know, big head in Silicon Valley on H B O and just, you know, rest, invest on the roof and, and, and chill for a couple years. But, you know, if, if you sort of embrace it in a way, it’s a, it’s a tremendous opportunity to like, take risks within the company itself. Like I, I think the way that’s what, you know, Teradata wanted me to do, there was a guy named Scott now who, who was really the principal sponsor of our acquisition. He, he, he later actually became c t o of, of Hortonworks. but, you know, he just gave me latitude and, and the courage to sort of like, you know, go disrupt some stuff, you know And, and so that’s, that’s what we tried to do, you know, while we were there. It was a lot of fun.
Dave Mariani: Yeah. again, overlap again. Justin Scott now is also very, very good friend of ours and a customer of AtScale actually at InterSystems. but, you know, you, you do need that sort of somebody within a big company. I, you know, I had, Scott, Scott Kaufman was my, was my guy. and, and so you need somebody so sort of within the company who can, who’s really willing to take risks in the big company and give you the space to do that. And so, and so, you know, that’s where the idea for AtScale came about, because I tried to do big things, ran into many, many, many problems, you know, trying to do that, those big things with big data, that, you know, and saw a space where there was a gap in the market. And, and so, you know, really Yahoo gave me the ability to sort of test out some of those ideas AtScale and take some risks. and so now we’re both taking greater risks and starting our own companies, but, but, you know, that’s, that’s the journey I think, of an entrepreneur.
Justin Borgman: Yeah. And I, I’m curious if you feel the same, but, doing it the second time around, like, you’re just so much wiser than the, than the first . Would you agree with that I mean, there’s just so much learning in, in the job that, you know, I, I feel like the difference between me as a, you know, founder and c e o 10 years ago versus today is just like almost two different people.
Dave Mariani: Yeah. You know, there’s, there is no, there is no substitute for experience. you know, and, you know, but there is like, you know, and, and, you know, in raising money at, especially at, at my age, you’re, you’re young, so you have no trouble with this, but it, there’s two sides to that coin, Justin. And, and especially in the venture industry, older entrepreneurs are, are, are really frowned upon. and they’re frowned upon because, you know, if you think about it as we get older in life, I know you just had another, another, another kid. Yeah. you got that mortgage, you got the family, and, and the idea is that you’re gonna take less risks as a result, is that you, because you, because you can’t afford to take those risks. Whereas if straight outta school, like you were, when you started adapt, is that, you know, basically, you know, all you guys as an department, no mortgage, you know, sky’s the limit.
Dave Mariani: No, probably no girlfriend, so, or boyfriend, whatever. So you can like, you know, work all hours of the day. And, and that’s, that’s true in some, in, in, in a large degree that’s very true. You, you have more energy and more time to put towards adventure. But like you said, it’s like, without that experience, without that gray hair, I’m seeing it, you’re gonna make first time mistakes during that first time, the, the, the first time at your event, new venture. And, and sometimes those mistakes, actually, most of the time, those, those mistakes can be fatal, to the company. So I think there’s a, there’s a lot to be said for having, some experience. and I don’t believe that because you’re older, you lack energy or lack the willingness to take risks. Yeah. I, yeah. But, anyway, that’s, that’s, that’s awesome. is there any other sort of, before we sort of move on to sort of more of the, the technical stuff that we’re doing, Justin, is there anything else you’d give advice you would give for an entrepreneur, either somebody who’s in it or somebody who’s thinking about, potentially wanting to start a company of their own
Justin Borgman: you know, yeah, I’ll give a piece of advice that I, that I also give to entrepreneurs. It, it’s sometimes considered controversial, so, so maybe that’ll be interesting for the viewers, although I don’t find it to be particularly controversial, which is that I strongly recommend, aspiring entrepreneurs quit whatever they’re doing, and pour themselves entirely into the idea of, of what they’re trying to build, rather than doing both at the same time. I meet a lot of people that are trying to like, hedge that risk by working their nine to five job or whatever their, their, their job is, to keep that income going and then get this thing going on the side. And I just think startups are too hard to be a hobby, too hard to be a side hustle. And, and, and the other thing related to that is I very much believe that the survival instinct that we have as animals here is a really important motivator when, when you’re building that business. Like, I think fear is a very healthy thing. And once you’ve kind of like really traded something off, you, you, you’ve, you’re, you’re, you’re very aware of the sacrifice that you’ve made, and that, that sort of, to me can be an even more powerful driver, like in a, in a very, you know, simplistic way for me leaving school, which my mom thought was like a terrible idea, of course
Dave Mariani: Mm-hmm. ,
Justin Borgman: you know, was like, holy crap, I better not screw this up like this. I can’t have this be a really bad decision. I can’t, you know, go back, you know, with my tail between my legs. I’ve gotta make this successful. And so that was always a driver that, you know, got me up early and got me, you know, working as hard as I possibly could. And so I think when you’ve got that comfortable job at the same time, it’s very hard to, to do both. So that’s my, you know, sometimes controversial piece of advice.
Dave Mariani: Yeah. It’s, it’s working without a net, isn’t it so, yeah, it’s a good, and so I’ll second that advice. you know, I left clout to start AtScale, and I was unemployed for a year as I wrote the business plan and, and developed the prototype. and then spent, went on at the fundraising march. and that’s, that just could not have been a part-time job. It’s, is not possible, like you said. Yep. So I think that’s fantastic advice. Okay. All right. Well, let’s talk about, let’s talk about the, the state of, the state of analytics and data. and, and, and, you know, Starburst is, you know, you hit the, you hit the scene with a, a, a, with a vengeance. I mean, it’s a pretty amazing how, you know, you were able to co-opt what was happening really at Facebook where, where Presto was invented, sort of incubate and build out a real business, you know, within Teradata, and then start Starburst, you know, with a, with a bang. and so, you know, where do you see, so how do you see customers and your customers using Starburst And literally what’s changed Justin in terms of how they use Starburst versus how they traditionally have, have, have run analytics against, databases and data warehouses and so forth Yeah,
Justin Borgman: So I think there are two things that I would credit towards. you know, why our value proposition seems to resonate with customers. I think, number one, I do believe there is a resurgence in, interest in data lakes as a, large center of gravity for your data. And when I say resurgence, I mean, like, again, if we go back a little over a decade, of course, Hadoop was the first data lake that was synonymous with the data lake. I think a lot of the principles of Hadoop, live on today, even though Hadoop, you know, is, is significantly less popular. We could, we could talk about why, why that is, but a lot of the concepts, live on the idea of open data formats, the idea of low cost commodity storage, the idea of something incredibly scalable. And in the cloud era, I think you are seeing more and more companies, customers, use data Lakes as a, as a huge foundational element of their data strategy.
Justin Borgman: And, you know, Starburst and, and Trino and, and Presto, you know, are, you know, is a world-class sequel engine for that. So if you wanna do data warehousing analytics within your data lake, we’re a great solution. And that’s really, again, you know, why Facebook built it, you know, Airbnb, Netflix, LinkedIn, a lot of the hyperscalers, leverage our technology, for those types of use cases. So that’s, that’s one piece of it. The second piece, and, and this is what really drew me to it while I was at Teradata, was that it recognizes a practical reality, which is that you’re not going to have everything in one place. You’re going to have data in multiple data sources. And every large company has this, this, this phenomenon. And that became really apparent to me when I was at Teradata because despite them being the industry leader in the enterprise data warehouse, this idea of like really bringing all of your data into one data warehouse, not one of their customers actually did that.
Justin Borgman: Every single customer had data that lived outside of Teradata. And that got me to think that, you know, that model’s actually impossible. I don’t actually think you can create a singular, monolithic enterprise data warehouse that has all of your data. If you’re beyond, you know, a, a, a significant size, maybe a small little startup can do that. But when you’re a larger company, that’s, that’s just impossible. Mm-hmm. . And so the idea of being able to federate and query across multiple data sources, extend beyond that data lake, was a big differentiator as well. And that’s, that’s really, I think what helped us, you know, get going. And, you know, in the early two years we were actually bootstrapped, we didn’t raise venture right away, which was also a fun way to get a business off the ground. really just working with existing users of the open source project and converting them into customers. And then from there, starting to really scale the business.
Dave Mariani: Yeah, that was sweet. That was so, I mean, man, that’s a dream to start a business without having to raise capital. and so you started the business making money, that’s amazing. I wanna go back to what you said here, a little bit about, about Hadoop. Cuz you know, we both got started in Hadoop. We both sort of bet a lot on Hadoop, obviously. Yeah. and, and, and obviously, you know, Hadoop is seen as a big failure. Yeah. So, talk a little, talk a little bit about that if you wouldn’t mind. you know, it’s like, cause, and, and then I’ll contrast it to what, cuz it was the first data lake, and Data Lakes, I’m completely with you. That data lake, that at Data Lake is a great pattern. especially if you can add engines on top of it to do the work that you need to do.
Dave Mariani: so if you have data in a common format, you can add the engine, whether it’s, an engine for data science or, or analytics, like you have with Starburst, you know, wide load it and reformat and copy data and store it in a proprietary format to do a particular job. If you don’t have to do that, that’s wonderful. Right Not having to copy data. So, so that all makes sense. And Hadoop, that was like one of the fundamental ideas about a Hadoop was to be able to store data in common formats, and be able to allow different operators on that data. So, so in your opinion, Justin, why did it fail
Justin Borgman: Yeah, I I attributed it to a couple things, and I’d love to hear your perspective on this too, since, since I’m sure you’ve reflected a lot on it over the past decade as well. to me, one of the, the, the big issues was that I think the hype, preceded the, the reality to some degree. Like, like it was such a big deal in that period of time. You know, again, 2000 10, 11, 12, and so much capital raised by the principle Hadu vendors, you know, Cloudera had raised over a billion dollars, which back then was utterly unheard of. Like, I mean, today, yeah, there are some companies that raise a billion dollars, but back then nobody did that. Nobody did that. I mean, that was like, you know, record breaking. and then Hortonworks was chasing them. and, you know, through all of that, I think that they, oversold in my opinion, what Hado could do.
Justin Borgman: you know, even though our vision was to, to make it the data warehouse of the future, and they came around to that vision and introduced actually a competing product, to us. so we would agree on the long-term vision. I think they oversold where it was on day one. And so you, you ended up with a lot of customers that felt like it didn’t live up to the promise. Like, who would deploy Hadoop, start to throw stuff into it, say, oh, cool, I can get rid of Teradata, run SQL queries on Hadoop. And they weren’t getting the performance that they needed. They weren’t able to do updates and deletes of their data. And so it was like, you know, this really, immature, you know, version of, of what people thought it could be. And so I, I think it left a lot of bad taste in people’s mouths, by not living up to that promise.
Justin Borgman: And, you know, there was a analyst who covered the database space for a long time named Kurt Monash, who I remember, you know, one of his posts, he, he said, it takes like six years to build a database. And it, and I, I think that’s right. It takes a really long time to build a real functional database and, you know, the, the query engines at that time, including Adapt, just hadn’t had enough, years behind it to really mature that. and so, you know, again, I think it was like kind of ahead of its time and promise too much, too early, too soon. And, and I think actually all those venture capital dollars are probably part of the reason why, right If you raise a billion dollars, you’ve gotta make this thing big fast. And I think that led, you know, some of those folks to, to be a little bit too aggressive. The other issue to me is that, and, and I hope I’m not offending any Cloudera alumni in the audience here, but I, I think despite their name, they missed the cloud era, and that was a mm-hmm. , that was a huge miss. yes. Cause the entire industry was moving in that direction and they just weren’t playing there. I mean, they, they’ve been so late to that game. So,
Dave Mariani: Yeah, I think, I, I would agree with you on what you said. You know, it’s, for, for, for us, you know, our first, our first sort of, data platform that the semantic layer talked to was Hadoop. And, and we sold our first, you know, our first dozen customers on Hadoop. And it, it became pretty apparent really fast that our customers were just struggling for making Hadoop work. Mm-hmm. and, you know, for, for AtScale, that was a, an existential crisis because, you know, because because we relied on the data platform to do the work, to do the, that that that last mile when it came to data, and so we couldn’t be successful if Hadoop wasn’t successful. And what, what I saw was that, you know, Hadoop made the mistake and the vendors, and they made the mistake of sort of like trying to do too much Yes.
Dave Mariani: and not, and doing too much poorly, right Yep. So it’s like, it’s an inch, it’s an inch deep and a mile wide. Yep. And so nothing really worked well, right It’s like, it’s like, yes, it could do job management, but it really didn’t do it well at all. Yes. Like, yes, you had a SQL interface, but it didn’t work really well at all. and so, you know, and so instead of sort of focusing on a few use cases and making those work really well, they kept on adding more and more features and more and more versions of packages. And then, you know, the, the big Hadoop vendors end up just being basically packages and they were just packaging open source and not really adding any value on top of it. so, so ultimately I think customers are like, like, like you said with the cloud, it’s like it’s, they, they’re, they, you know, you know, home Depot is not in the business of, of, of operating their own data platform, right.
Dave Mariani: And they’re in the business to sell products, for home improvement. And so, so what I saw is that, you know, customers experimented, it wasn’t successful, and then when the cloud was ready for them, they saw the cloud as a way out of letting somebody else manage that for them and them getting back to business. Yeah. and but, but the ideas, like you said, of the data lake are I think really starting to ex starting to be realized finally. Yeah. And what, what’s that It’s, it was, it’s almost 15 years later, right
Justin Borgman: Right. Yeah, totally. yeah, hundred percent agree. I I think you made a good point in there too, for any aspiring entrepreneurs, which is, don’t try to do everything, like focus on being the best at one thing, you know And I, I think just as a startup, you don’t have enough resources, even if you do raise a billion dollars to do everything incredibly well. So I think there’s less in there too.
Dave Mariani: Definitely Lessing there. So, so, hey, Justin, when it comes to sort of how you’re seeing, customers deploy Starburst or just what you’re seeing in the market in general, there’s this whole idea of decentralizing analytics. you know, where, you know, we were born, oh, sorry. I’m just saying, I was born of the era where , where everything was done by a central data team, and the central data team did everything data pipelines. They also, you know, produced, you know, the bi platforms of the days, the business objects and the like, they, they built those cubes. They may have even built the dashboards and, and the reports for the business. And, you know, and then we saw the tableau come on the scene and just blow that up so that the business was doing everything themselves. And now we’re sort of like, I don’t know, maybe somewhere in the middle, but people talk about data mesh. I like to talk, talk about it as hub and spoke. What are you seeing out there with this sort of new trend
Justin Borgman: Yeah. I, I think it’s very interesting, but I also be, before I share some comments, I will caution, you know, the audience that like any trend, these things do not happen from zero to one overnight. Right. And so, don’t expect that everybody deploys a data mesh tomorrow. I, I don’t think that’s what will happen here. I think instead it’s, more evolutionary. And I think it depends on the organization in terms of whether it makes sense for you. I think central teams will always play an important role for things that logically should be centralized, you know, governance and things of that nature. But what mm-hmm. data mesh really represents, I think, is an acknowledgement that data is inherently decentralized in large organizations. And how to really turn that into a strength. Like, how can we empower the, the data domain owners, the people who know the data the best to play a more active role in, you know, data management and how, how do we think about treating data as a, a first class product that we make available in this organization
Justin Borgman: And I think a lot of those concepts are very interesting. you know, the notion of hub and spoke, I like that term. That’s actually what we use when we’re training our, our sales reps as a matter of fact, because they’re both important and I think mm-hmm. , you have to be aware of that, right Like, central data teams will play a really important role, but you also have to get buy-in from the spokes. And, and I think it’s that collaboration between those two that creates, you know, success for customers and, and, you know, hopefully success for vendors as well that, that can help facilitate that.
Dave Mariani: Yeah. You know, me being in, when I was at Yahoo and running analytics, it’s like, it’s, it was like, there’s just no way I could have a team understand every segment of our business. Yeah. In our case, it was a advertising and, and, and, and our audience, it’s like, look, it’s like the business understands their problem and the DA and what they need to solve their problem. Yeah. And the kind of data they need to solve the problem. So if we can enable them, that’s fantastic, but you can’t just let, let, you can’t just turn them into it. People, that’s not what they want, and that’s not, that’s not their job either. So you gotta have some kind of help, and standards. And so I do like the hub and spoke. I use the same thing, when talking about, talking about just how you decentralize and, and, and leverage more people in the organization to do analytics.
Dave Mariani: Yeah. so let, let, let, let, so let’s pivot a little bit, and talk about costs as this whole idea, fops is, is a, is a new term that’s sort of being, being, being talked about a lot because people have moved from on-prem where we went from sort of a CapEx right And into the cloud where now it’s opex. Yeah. And now in your opex, everything matters. So what are you seeing with your customers What are you seeing in the market, in terms of how people are dealing with this sort of new paradigm where, you know, they’re, they’re running analytics in opex without CapEx or, or, you know, instead of CapEx
Justin Borgman: Yeah. I, I think it’s, it’s important and probably particularly important right now with some macroeconomic uncertainty. Will there be a recession You know, people are more focused on those operating costs than they have been maybe for the past couple years. And, you know, simultaneously people are recognizing that consumption models, while, while I think there are a ton of benefits to consumption models and, and we generally sell as a consumption model as well, you, you have to keep an eye on them, right Or they, they can, they can get outta control. And, and there’s an element of this that is to me, also replaying history. Like, I think if we go back, you know, 40, 40 years of database history, with the early advent of the enterprise data warehouse model, you would consolidate data in, in one place, maybe that was Teradata, you know, my former employer.
Justin Borgman: but in the process, you’re locking your data into the hands of one single vendor, and those costs generally start to, to increase and you feel locked in. It’s very hard to sort of get out of those situations where you put your most valuable data in the hands of one vendor in a, in a proprietary data format. So I think that movie, which, you know, led to some resentment towards, you know, Teradata, Oracle, some, some of those vendors is now replaying itself in the cloud world, with, you know, snowflake being the, the, the best example of that, obviously the most successful cloud data warehouse today. but that success also comes with spiraling costs for, for customers, at least those that have achieved some level of scale. And so I think it’s, you know, customers now are starting to think about how do they rationalize that spend
Justin Borgman: How do they manage that spend And really, does all data need to fly first class in an expensive cloud data warehouse, or does some of that data work in a data lake You know, again, it’s kind of back to a lot of those concepts of, you know, 15 years ago where it was Teradata and Hadoop, well, maybe now I should have Snowflake and, and S3 Data Lake, you know, and I should, should use Iceberg or something as a data format, which is open source, and I can use multiple engines to query it. So I think that’s what you’re seeing, and I think that’s also what’s driving, driving some of the increased attention in data lake models now as well.
Dave Mariani: Yeah. It’s like a, you know, do you really need that query every hour or, or every minute, or can you get it once a day I mean, it’s like a, you know, you have to start a asking yourselves those questions because you know, it, it costs money for, for, you know, lower latency. It cost money for, you know, like for to fly first class. I like that. I, I like, I like your analogy. That’s an awesome way of thinking about it. Okay, so let’s, let’s just close it out. I’ve had you for way too long. but for something that’s a little bit, a little more fun, lots of, there’s, there’s a lot of talk about G B T and, and, and how, and, and large language models and how that might affect analytics. Yeah. what’s your, what’s your perspective on that, Justin
Justin Borgman: I think tremendous future. Like, I’m, I’m very bullish on the potential, but I think we’re in like the earliest innings, right Like right now, I, I, I’m sure this is the case in, in your company as well, like, you know, engineers are playing with these tools to write poems, you know, about, you know, or ballads or, or epic stories, you know, about, you know, random things, like people are having fun with it, right It’s sort of almost like a, a, a gimmick or a curiosity, but I, I think that undermines its potential, which to me, especially in the world of data, is pretty interesting. Like, if you could, automatically craft valuable data products that, that your business users, you know, need and, and can get value from, that’s hugely powerful. you know, I, I think it, it, it potentially could play some disruptive role in the, in the bi landscape to, to some degree as well. so I think there’s a lot of potential, but we’re just barely scratching the surface. And, I think there are likely going to be, if not already, a number of interesting, I think startups that will be founded around, you know, how you leverage generative AI to help facilitate or automate, you know, elements of, of analytics going forward.
Dave Mariani: I like it. Well, let’s, like, let’s hope those startups don’t raise too much money, Justin and , and try and try to do too much . That’s right. Well, hey, Justin was great hanging out with you. you have a great background and, and you built a, a, a large company, you know, on a shoer. So you’re, you’re, you, you’re, you’re a role model for a lot of people out there, who are looking to, build their own businesses. So I wanna thank you for hanging out with me today. Thank you. And, and, and thanks all the listeners for, hanging out with us. So thanks everybody.
Justin Borgman: Take care.