The Data Stack Show - The PRQL: Managing Complexities of Financial Data

Episode Date: November 28, 2022

In this bonus episode, Eric and Kostas preview their upcoming conversation with Ashwin Kamath of Spectre.

Welcome to the Data Stack Show prequel, where we give you a preview of the show that we just recorded. Boy, this one was fascinating. My question for you, based on this show, Kostas, is do you think that it's possible for you and I to go raise money to start a hedge fund? Oh, yeah, absolutely. Yeah.

Yeah, yeah. Yeah. Yeah. Yeah. I think we should add like a new, a new series of webisodes that it's going to be like our experience of trying to start the hedge funds. Hmm. Let's do it. Let's do that. I think we could have a GoFundMe campaign for like initial.

I mean, I don't know. It's like how you start the hedge fund. I have no idea. I don't think so. Maybe we should try it. That's true. I think there is a little bit of a difference between like personal donations and institutional capital. Okay.

But Ashwin, who we talked with on the show in all seriousness so he worked for a firm which is like a they provided sort of call them like micro loans like one of the first big ones in the market and then he went to work for one of the most like technologically advanced hedge funds in the entire world and built out a ton of their data infrastructure. And I was blown away at how much third-party data, as he called it, or they call it alternative data in the finance industry, which was interesting to me, a new term for me personally. I was blown away at how much data, I mean, terabytes of that data, right? I mean, we sort of traffic in a lot of first-party data and sort of our work and the customers that we serve at our companies.

But goodness gracious, terabytes of alternative third-party data is crazy. And it has to be accurate because they're doing things like algorithms that will like longer short of stock right so if you get it wrong it's really really bad uh which was crazy so i mean what did you think yeah first of all i i just remembered something that's probably like a very like probably everyone knows orn Haskell, like this company. And it's like, it's like the business model is exactly that. Like they are like selling data and that's for Square, right? I don't know how many people remember for Square, but the whole idea of.

David Pérez- Like check-in, but then they actually were just a data brokerage and then they're selling the check-in data. Misha Milovskii, Yeah. So they were gamifying, let's say the whole like moving around and like being a mayor and whatever. And yeah, like this company with the data that has collected like in this past like decade or something is like fueling like some pretty big applications that they need toolocated data

out there. Sure. So that's what they're doing. They're selling like third-party data, right? Geolocation data. Yeah. So it is out there. I think it's just like something that we don't, as a consumer, you don't really get exposure

to it. Like, I mean, you are exposed like to the results of using this data, but you are not aware of it, right? Yeah, fraud detection or getting like a real-time approval for a loan, right? But all these services are based on, I don't know how many different signals that are collected, not just by the company, because the financial institution that gives the loan, grants the loan, doesn't necessarily have access or can generate this data, right?

So there is an interesting dimension, the data market out there has to do with that. And we haven't talked that much about it. I think it's worth investing a little bit more in like figuring out

and like sharing. Yeah, it'd actually be interesting, Brooks. We should try to find someone who is a broker in that space and get them on the show. That would be

super interesting. The other really interesting thing, which we're probably over time because we've been a little bit verbose today, but the other really interesting thing was that Ashwin really, he framed data quality in terms of practically in terms of orchestration, which was really interesting to me. I mean, he talked about monitoring, but he didn't discuss monitoring in terms of the hot companies out there that are doing data observability or whatever, right? He really, I mean, he was multiple times brought up data quality as a function of requirements in orchestration.

And that to me was also really interesting. So if you want to hear about that, definitely check the episode out. Absolutely. All right. And with that, this has the episode out. Absolutely. All right. And with that, this has been the Data Sack Show prequel. We'll catch you on the next one. Subscribe if you haven't.

We also have a newsletter. We haven't mentioned this in a long time. If you subscribe, let us know. We also have a bunch of swag. If you want swag, just reach out. All you have to do is give us some feedback on what you like and don't like about the show. You can email us, fill out the form on the site, et cetera. And Brooks will package up a shirt and a mug for you personally with a handwritten note, and we'll get on the way.

Catch you on the next one.