The Data Stack Show - Shop Talk: The Business of Data Infrastructure is Uniquely Challenging

Episode Date: November 4, 2022

In this bonus episode, Eric and Kostas talk shop surrounding data infrastructure.

Welcome to the Data Sack Show, Shop.Costas. We have talked with people who built amazing data technology at companies like Netflix, Uber, and LinkedIn. But you and I actually don't record our talks about data very much. But we actually talk about data together a ton. And so Brooks had this amazing idea of just recording some of the conversations that you and I have before and after the show about data and our opinions on it. And really, this has been one of my favorite things that we do. So welcome to Shop Talk. It is where Costas and I share opinions and thoughts on a personal level about what we're seeing in

the data space. And it really is simple. We ask one another a question and the other one tries to answer it. So without further ado, here is Shop Talk. Welcome to the Data Stack Show Shop Talk, where Costas and I talk back and forth about the topics that interest us. And we've learned a ton. These are great. I've learned a ton in these talks. And I believe it's your turn today, Kostas. So what's been on your mind? Kostas Pintasenac Yeah.

So I have a question for you. You've been at Thunderstock like from very early stage of the company, right? Yep. But this is not your first startup, but I think it's your first startup that has a built data infrastructure, right? Right. So my question, and now that you have been like for a while building a business in this industry, I'd like to hear from you, like, what are the differences that you see between like building a business around the product, like in the data infrastructure compared to your previous experiences. And I'm asking that because, okay, but I've been working like for quite a while, like with startups, but pretty much all of them have become like data. So I never had like, let's say the opportunity to experience like how it is to be like a technology company again, but like something different. So tell me, how does it feel? Is it different?

Is it better? Or... David Pérez- That's a great question. Maybe, would it be helpful if I described previous companies? So there's like a baseline. I won't spend too much time, but maybe that'll be helpful context. So the first company was

an education startup. Specifically, it was actually for-profit education and what people today know as like code boot camps. And, you know, there's I haven't looked at the industry in a while, but there was kind of like this big bubble at a point with like, you know, people learning to code, demand for software developers or whatever.

And that was, I think, in like 2011, 2012. So actually back then, there were really only a few like code boot camps, as people call them today, like one in New York and like I think a couple in San Francisco. And that started out as a very traditional business. Like it was for-profit education, teaching people how to code, and then placing them in jobs. And we actually started out with all physical locations. So at one point we had like 30 campuses or something like that. So it became like a pretty large business. And then of course we were working on digitizing it and productizing it.

So we actually built essentially an LMS. Bought this like a standalone product as an acquisition fee for the in-person product. And then also to like facilitate the in-person product so it was really interesting and then ended up selling that to a large publicly traded company and then you know whatever the other didn't whatever there have been a couple of like small like light whatever which what the venture you know world would call like a lifestyle business like a consultancy or whatever just a basic cash flow business and then in terms of like technology startups there's also actually i don't know if i've ever told you about this we built a technology that allowed you to like take your retargeting audiences from your website and essentially resell them to other to like advertisers that sounds very evil to me

like I mean yeah it's like yeah totally to you it would sound it would sound totally evil for a marketer it's a dream right if I could like serve like people visit

your website and then they go out and browse other websites. Normally you would retarget them with your own ads, but it would allow me to retarget your audience with my ads. But you know, based on that stuff, which are certain businesses is very powerful, right? So like if you have two non-competitive businesses, it's a way for them to create a ton of value without having to like share data. So in many ways, actually, one of the big use cases was actually protecting privacy.

Now, possibility for abuse, significant. But used correctly actually was really cool. And COVID basically decimated that business, right? Like when COVID hit, everyone basically just completely stopped spending on anything exploratory from an advertising standpoint it just essentially killed the business so okay so what's different so one thing i would say is that the The marketing is way harder.

And I don't say that to diminish. And really, the main reference point is, well, in both of those cases, we can compare both examples, right? I'm not diminishing. Consumer marketing is very hard, right? Especially if you think about consumer mobile or, you know, consumer mobile games or whatever,

it is phenomenally difficult. So I'm not diminishing the difficulty. It's just difficult. Like it's a different type of difficulty. So for example, the main marketing message for the education startup

was learn to code, get a job, right? I mean, it doesn't get more simple than that. And that's also like an extremely powerful value proposition, right? And, you know, whatever, like that business happened to like provide a very relevant product at a relevant time and like the message really resonated and it was like very simple right and even with the other one you like you talk to a marketer and it's like well would you like to serve my retargeting audience ads and they're like yes i would right yeah i'm a marketer i will follow everyone everywhere when you talk about

when you think about data infrastructure the nature of the problem is in reality a lot more complex it just is right like the like, what does this infrastructure do? How does that fit into the larger picture of like technology within a company? How is it differentiated from like other infrastructure? The details there are, it just takes a lot. It's all, I would, in some ways, trying to think of the right word here, because I don't want to diminish like how hard consumer is. Cause that like, especially the brand building side of that was really hard,

but this is what I would say. One of the major things that I've noticed is that the, there's a lot more requirement to take a very high level of complexity and distill it down. And that process of distillation is very difficult, not only from the standpoint of how do we describe our own product, but also like, how do we differentiate it? Right. And like a great example is, okay,

you have like ETL, like this is your space, right? Like you had ETL vendors and it's like, well, how is your ETL pipeline different than this one? Right. And it's like,

well, I mean, meaningfully communicating about the differences without bogging people down in technical details that aren't necessarily helpful to explain how it can help them is really hard. That's really hard. So I don't know. That's the first thing that came to mind. Yeah. Yeah.

Yeah, that's to mind. Sergio Leal Yeah. Yeah. Yeah. That's, that's super interesting. And I have felt that too, when I was trying to, to build like my first business, because like, you know, like when you're looking for marketing, you have marketing advice and you have like zero marketing experience, for example, like all the advice that you will get from people usually comes like from consumer-related marketing, right? So yeah, like you will get advice about doing stuff that sounds very reasonable, but probably they're like super hard to do because you have to distill some very

abstract concepts into something tangible. Like, yeah, learn to code, get a joke. That's great. Amazing. Yeah, sure. But how do you phrase and communicate the same, let's say, amount of information for building a pipeline and moving the data around. So yeah, I feel you. And okay, so if you have to choose your next startup to work at. Ooh, like an existing one or just... I mean, it's not about like the startup itself. It's more about like, let's say the industry. You went like from Med tech to like data infrastructure.

What do you Alex Wrigley- With the tour detour and evil advertising technology. Yeah. Like we will keep this like, you know, like under the radar because we love you and okay, we all make mistakes, you know, it's like, it's okay. David Pérez- I definitely would never do ad tech again. I can tell you that. What do you want to do next?

Is there like any industry out there? Like anything, like for example, like, okay, like the data industry is huge, right? Like someone who's working like in ML Ops is like probably like, it's not something that I would like easily do. Like, I don't know enough to go and do it, although I'm working like in the data industry, right? But regardless of like how much you know from all the things that you have seen

so far, like all the stuff that you have been exposed to because of the show so far, right, like part of the beauty of this show is like getting in touch and thinking about like stuff that we wouldn't otherwise. What excites you? Like what you would do next? David Pérez- This may not have been part of the question, but I think that I would definitely stay in the data industry.

Um, part just because I love, I love the technology side of things. And I mean, whatever you can argue that like all technology is data related. But, you know, I think obviously you and our listeners are talking about, you know, the types of stuff we have on the show. And it's really, I mean, it really is fascinating. main reasons I would stay involved is because I think we are in what we will look back on as one of the very exciting times. I mean, there are tons of exciting times ahead, but there's just a lot of new frontier being developed right in front of our eyes. So what an exciting time to work in this general space. Okay, I'm just going to, I haven't thought a lot about this. So I'm just going to tell you what first came to mind.

So there are two things that first came to mind. One is more of like a category or like a problem, a type of problem that I think could apply to like multiple different data technologies. And I'll explain what I mean by that. The problem type is the automation of things that are currently still pretty manual. So a couple of examples of that that come to mind are things like data quality, you know, or whatever. There are a number of terms there, right? Like data quality, observability, blah, blah, blah, right? But messy data is pervasive.

And at least as far as I can can see even though there's really cool technologies sort of emerging around it it's still a very widespread problem and i don't think that anyone has really figured out how to solve it in a great way i'm not saying that i'm the person to figure that out but i'm saying like as a type of problem, I'm very intrigued. This is probably a better way to say it. I'm very intrigued at data-related technologies that free up mind space for people working with data

to solve harder problems, provide more value, etc. Right? Like there's sort of, it seems like there are still a significant number of like fairly low level problems

that require an unnecessary amount of, you know, sort of call it like manual labor. So that's really interesting. The other one, like this is, this may sound crazy because there's no way that I'm qualified to do this at all, but you actually mentioned MLOps. The reason that is interesting to me is because every time it comes up on the show, you can see like the searing pain of difficulty that is MLOps right now, right? And that's not because like it's an ignored problem, right? I mean, we've talked with some amazingly brilliant people trying to solve this problem.

And so maybe this is a little bit masochistic, but I like really challenging problems for some reason. And so that, you know, that to me is, is really intriguing because the pain is so palpable, like when you talk about it, you know, so it just seems like this huge problem space, but, you know, I don't know anything about ML ops, so that would be a steep learning curve. I don't think that you would have like a big problem, to be honest. I'm pretty sure that you would be like super successful in doing it. And any company that would have you would be really, really blessed. Oh, that's good.

That is so flattering of you to say. Yeah. But if you are a recruiter, don't reach out to him. He has like very serious work to do. David Pérez- No, that's a great question though. It is really interesting to think about that. Okay. I think we're, we're really close to the buzzer, but I have to ask you what,

what problem would you go? Stig Brodersen I don't know. I think I think I would stay like in like storage and database, like space. I think there are like plenty of opportunities and challenges right now. So I would probably do that, but yeah, anything in Mandalorian is very fascinating because I just haven't done it. And I'd love to do something new and more. It feels more exotic. That's right. Yeah.

Like the Miami of the data world. I thought that this is the crypto. No, crypto is the Miami. Oh, crypto. Web3. Yeah. What's the Vegas of the data world? The Vegas of the data world.

Where should I go? It needs to be something that's like pretty extreme pretty expensive and like doesn't really result in anything except maybe you like wake up and you discovered you're married yeah Oracle

Oracle is you wake up in a marriage that you realize you signed a 10 year contract for and you can't get out of yeah it's a marriage that you realize you signed a 10-year contract for and you can't get out of. Yeah. I said Oracle because I think it's like this thing that you can't have until you realize the consequences of your choices and that you are legally bound to that for a long time. Oracle as the Vegas of the data space.

You know, Costas, we learn so much from the data leaders that we talk to, but I learn so much from picking your brain. And actually, your questions really make me think really hard. So I appreciate Shop Talk. I think it makes me a sharper thinker. Well, it's fun. Like, I think it's good to just sit and chat about the stuff that we experience. And yeah, I think like, I hope like people enjoy it.

That's why I'll keep asking for people to reach out. Please do this. Come on, folks. Like, you can do that. Like, send an email. Yeah. Let us know how you feel and, like,

what are your opinions of, like, your experience with the show. So, please do that so me and Derek, we can keep being happy. Please. Of course.

And, of course, we try to take the same types of questions to, you know, data leaders from all sorts of companies, large and small. So definitely subscribe to the main show if you haven't yet. Tons of really good episodes there and tons of really good thoughts from data leaders, you know, really around the world. So definitely subscribe if you haven't. And we'll catch you on the next Shop Talk.