The Data Stack Show - 240: Data Council Insights from a Waymo: Postgres and the Future of the Data Stack

Episode Date: May 7, 2025

Highlights from this week’s conversation include:Recording from a Waymo (0:54)Future of Data Technology (2:45)AI Integration in Data Work (4:20)Speeding Up Data Experiences (5:29)Snapshot Conversations with Founders (9:52)Diversity of Perspectives on Postgres (12:37)Cultural Significance of Database Mascots (14:09)Incubation and Success of Open Source Projects (16:43)Final Thoughts and Takeaways (17:34)The Data Stack Show is a weekly podcast powered by RudderStack, the CDP for developers. Each week we’ll talk to data engineers, analysts, and data scientists about their experience around building and maintaining data infrastructure, delivering data and data products, and driving better outcomes across their businesses with data.RudderStack helps businesses make the most out of their customer data while ensuring data privacy and security. To learn more about RudderStack visit rudderstack.com.

Hi, I'm Eric Dotz. And I'm John Wessel. Welcome to the Data Stack Show. The Data Stack Show is a podcast where we talk about the technical, business, and human challenges involved in data work. Join our casual conversations with innovators and data professionals to learn about new data technologies and how data teams are run at top companies. How to Create a Data Team with RutterSack Before we dig into today's episode,

we want to give a huge thanks to our presenting sponsor, RutterSack. They give us the equipment and time to do this show week in, week out, and provide you the valuable content. RutterSack provides customer data infrastructure and is used by the world's most innovative companies to collect, transform, and deliver their event data wherever it's needed, all

in real time. You can learn more at ruddersack.com. All right, this is the first episode of the Data Stack Show ever recorded. I guess live isn't really the right word, but ever recorded from a moving Waymo. We are being driven by robots here in San Francisco, riding around with no driver. We just got finished at data council this week in Oakland, recorded six, seven, six, six episodes. A lot of takeaways from the conference this year.

I think one thing that is abundantly clear to me is we are still figuring out how AI is changing everything, but it is changing everything. John, what are you thinking? Yeah, Brooks, a great conference and I was just sharing that one of my favorite parts about the conference was doing the show at the conference. Like we got to sit down with so many great people all in the same space in real life, which is rare for us. Very special.

And yeah, really special and got to go really deep with people on technology, on startup life, you know, a lot of different things. But as far as high level themes, I think one of the slides that came up was a description like trying to explain the evolution of the data stack. And it talked like, you know, we were here and it like kind of described the like, I don't know, legacy era, whatever you want to call it, historical era where it was like Oracle and SQL Server and you know, and that was like a data warehouse.

That era like evolved all the way to current, to kind of where we are now past the modern data stack, the Databricks, you know, Redshift's part of that Snowflake era, et cetera. And the interesting part is they had a big question. They had a break around 2024 and then a question mark on the screen. And like this is like one of the premier conferences like we're here to like fill in that question mark. And then the words behind it were like,

they didn't just leave it at a question mark. It was essentially like iceberg question mark. Like we think we feel pretty confident iceberg is part of this and the rest of it's a little bit still TBD. So it was a unique year to go. Yeah, absolutely. Absolutely. Yeah, I mean, one thing I've been kind of thinking about too

is I don't want to tell them talk about the modern data stack but one of the panels in particular, I think the modern data stack was part of one of the titles or something, and we all joked about this in 2021, 22. I can't remember when the whole modern data stack was just like, the hype was kind of fever pitch, and some people started saying, well, you know, what is it, one day it's gonna be the post-modern data stack?

And I think that was kind of some of my thoughts that this week is like, I mean, now is that now, which we refer to that as like the traditional data stack. I don't know. Yeah, that's a great question. But clearly it is, it is changing. You can't see us with Brooks and I are both like checking both directions.

As the Waymo makes a turn here. Makes a turn, there's cars parked on the street and it's like the visibility is like a little tricky. So yeah, wish us luck. Here we go. Purging and taking a right turn in the traffic. All right. We made it.

Pretty good execution. Yeah, so, you know, Iceberg was an obvious theme. And even like an evolution of like the AI being more incorporated into tracks as part of the conference and then obviously more incorporated into the data stack. I think one of the one of the other things like back to that like kind of big question mark theme who was talking about like found like founders like in this environment yeah we're essentially like there is a high amount of uncertainty. Like, well, will my company become a feature that like for like one of the big data companies,

we're for like one of the AI companies. Yeah. So I think that was another theme. But but one and one other really cool thing that we saw was people focused on speed of like, hey, like, how can we make this like experience faster? DuckDB had some neat things around that. They actually dropped a release, I think today, called Instant SQL, Mother Duck being like, you know, the company there, that

that was really neat. And I think like data practitioners are used to this waiting pattern of essentially like, hey, I'm gonna run the query, and then like I've got a couple seconds, I'm gonna. And then like, I've had a couple of seconds. I'm going to run the query again. And I've got a couple of seconds. And, and then essentially developing a way to sample the data. And then you essentially get like millisecond responses, like as you tweak the query. I think if that becomes like, obviously they're doing it.

I think others are going to do it too. Like that would actually fundamentally change the day-to-day of an analyst Yeah, and then the way it intersects with AI that that was interesting is they showed like a way where it's like hey Like you know prompt like essentially like type in a you know window and prompt a change to a query for example or a new query and That was really interesting too because the prompt you you know, the prompt modifies the SQL, but then it runs instantly. So you get instant feedback for what the prompt did visually.

And then after that, you get to like say, hey, I accept that change. Like I didn't like that change. I want to make another attempt. So I think those are some interesting patterns that like I expect will continue. And it's honestly surprising if you think about a typical workflow, something that's working with data,

there's not actually that many spots to have a fundamental shift of how you might work. And I think that's a real one. It's probably going to be a thing. Yep, totally. One thing I heard, I think it was the opening key, I think it was the opening keynote, and I was on a call, I think maybe walking into the auditorium or finishing something up.

So I didn't make the whole talking as I was out in the, I guess, atrium area, heard the speaker say something about the chat interface. And basically it was like, we're tired of the chat interface, it it's like we're you know we're tired of the chat interface it's not that great and everyone in the room just cheered and I just kind of chuckled to myself what are your thoughts on that John? Yeah I was in the room for that talk and it was really funny it was kind of a hot take on right maybe not even a hot take actually I think it was really well thought through to be honest but the take of like like hey like just because we have this chat interface

Doesn't mean it's right everywhere and like we should all be rushing to implement chat Yeah, everywhere which like before AI we'd never would have done We never would have like oh like we need an interment of chat interface here and here and here like nobody like nobody asked for this Like nobody wanted this Yeah, and it was a good like yeah, everybody cheered and it was a good like Moment to realize like okay cool. We've got a lot of people that are like developing these products they're gonna really try to think deeply about

Like how to use AI for sure but like also think deeply about the implementation and not just default to chat. That was cool to see. There's a lot of founders in the room like that is top of mind for people. We're going to take a quick break from the episode to talk about our sponsor, Rutter Stack. Now I could say a bunch of nice things as if I found a fancy new tool, but John has been implementing Rutter Stack for over half a decade.

John, you work with customer event data every day and you know how hard it can be to make But John has been implementing RutterStack We can collect and standardize data from anywhere, web, mobile, even server-side, and then send it to our downstream tools. Now, rumor has it that you have implemented the longest running production instance of RutterStack at six years and going. Yes, I can confirm that. And one of the reasons we picked RutterStack was that it does not store the data, and we can live stream data to our downstream tools. confirm that. set. Yeah, and even with technical tools, Eric, things like Kafka or PubSub, but you don't have to have all that complicated customer data infrastructure. Well, if you need to stream clean customer data to your entire stack, including your data infrastructure tools,

head over to rudderstack.com to learn more. So we talk with six folks in person at the conference recorded episodes. What are what were some of your favorite moments or kind of key takeaways from that, John? Yeah, so I think the fascinating thing for me personally is we got to just get snapshots and time from, I think we just talked to founders, essentially, trying to think founders are like, or yeah yeah or people that are like uh you know on an executive team like right with a founder right but

um but we got all these really cool snapshots in time like we talked to one company that was eight months in really like pre-product market fit I just have to give you guys a way little update really quick We just we're on a really packed street cars on both sides and we had a car coming and there clearly wasn't enough room for two cars to like come through this narrow street. The way I'm gonna like pulls over like let's the other car come through and then like just continuous. I mean totally handled it probably better than I would. Yeah.

Truly impressive. Yeah. Yeah. I'd own it probably better than I would. Yeah, 100%. That was truly impressive. Yeah, yeah. But, so the six founders, and it's so fun to be part of these moments of six snapshots. One that was pre-product market fit, eight months in, first time founder.

Yep. Which was a really cool conversation. We're excited to share that with you. Another seasoned, like, seasoned, not founder, but a seasoned executive, like it's got a bunch of companies and like done this a lot. And then some others in between some others are like kind of five years in or whatever. And it was just really cool like having that having a lot of those like back to back and they all really had like great things to say. It didn't like it didn't feel like you might think it would like, oh, like, man,

like we talked to that season founder is like, wow, like just like great stuff for sure. But even though like these founders that are like first time, like also had like some really like deep insights and great like things to say. So it was neat to see like a whole spectrum there. But but really not like this like progression that we think of as like this like junior to senior spectrum type of thing or junior to like highly experienced, but just like differently experienced.

Yeah. So I thought that was neat. Yeah, I know it was really cool. I mean, man, just a lot of really smart people solving really hard problems in very innovative ways. At a time where it's you know the ground is kind of changing beneath them yeah and which we discovered a bedrock thing though which was surprised to me

of like also almost all six not quite all six like post-gres is like the unchanging thing this year. We talked about Postgres in five of the six. Five of the six. Yeah, I think that's right. I mean, totally crazy. But yeah, I mean, it was kind of the constant and... That was a shocker for me. Going into it, I would have thought, oh, like this con, you know, there's all this innovation going on and we're going to talk about this and talk about this.

But to like do these interviews, and this was even was even true I think of the conference in general, but especially with our interviews of essentially like different perspectives on Postgres. Like keep everything in Postgres and like extend it. Like that's a good move. We're like use Postgres, but like stream the data out of it. This other thing. That's true. And like, but nobody like said don't use Postgres.

That was not a take of anyone. Yeah, that's right. Like, I mean, obviously some of the takes were like don't use Postgres. That was not a take of anyone. Yeah, that's right. Like, I mean, obviously some of the takes were like, don't use Postgres for everything. Yeah. But yeah, I think that was a little bit unexpected, especially for the, you know, the kind of the age of the technology

and there's so many like other things, like, you know what we didn't talk about? Didn't talk about MySQL. That didn't really come up. We didn't talk about like, you know, like some other like things that, that like 10 or 15 years ago, you're like, oh well, yeah, this is here to stay. Yeah. So it's just, it's interesting what stays and what goes.

Yeah, I mean, so interesting to see that, you know, there's just kind of, well, everything is changing kind of an anchor or a constant. As people are figuring out What are the other pieces gonna look like it seems it's almost been like a helpful thing for folks to Find this piece that there seems to be a lot of agreement You know that this will stay and be a part of whatever things around that look like is really cool. You know what I just thought of? Do you know that like, I don't know, mascot is probably not a fair, the mascot for

Postgres? No. The animal. Oh yeah. So the animal, like a lot of listeners know the animal is an elephant. Oh, you know what I'm saying? Yeah, that's right. That's like a really interesting thing for like, hey, like this is the end in my,

I'm going to be embarrassed if I get this wrong, but like my sequel, I'm pretty sure is like a dolphin. Yeah. It was like the animal that kind of went with it. I think that's right. But like interesting, like the one that's like really like has stuck around and like, like elephants are good. Like that was a good choice.

Like, and we had this conversation with one of the founders about like, about like this long-term thinking and about like, about the culture that's set. Yeah. And the expectations are set. So like, somebody, like, there might be a correlation here between like, whoever started, like, whoever is like, influential, like, you know, in early days and thinking this and like, the longevity. I don't know. Yeah Yeah some foresight to picking the elephant. Yeah. What was it? It reminded me of talking about Clickhouse. The name basically Clickstream Data Warehouse. Yes. Can't remember exactly what Aaron's pointed around that was but I started as an open source project but

they named it Clickhouse purposefully kind of knowing that it would be something different one day. Right. It was pretty cool. Yeah. Yeah, there's this Well, that was a fun topic too with Aaron of talking about like hey like somebody asked so he did a talk Aaron from ClickHouse like did a talk and it came up on this talk and on the podcast of like Should I open source should I start this thing in open source? And Aaron's is like no Don't do that Which is this a funny like funny thing And it makes a lot of sense. And, and, but I think, but I love that. But that's so helpful, I think to people that are in that stage, like

trying to start something like, should I do open source or not? And no is a great answer actually in both ways of like, no, like, oh yeah, okay, I'm not going to do that. And like, no, nah, I gotta do it. Like, you know what I mean? Like like oh, yeah, okay. I'm not gonna do that and like no not I gotta do it like you know What do you mean? It's just a helpful answer Well, I mean it's interesting. I think maybe the pragmatic take is no But at the same time he's talking about click house and was telling us there's no way the technology would be where it is without the innovation of all of the open source contributors and people just constantly working on this technology, evolving this technology.

And I think it's similar to Postgres, right? It would not be what it is unless it's open source. Yeah, I mean, it's just super interesting to think about. Well, and I think the model, and I don't think anybody said this explicitly this weekend, but if you, me thinking through these projects that have been like really successful, they incubated inside of companies, they incubated inside of universities or like in the case of DBT incubated inside of an agency actually.

So like these long incubation times without burning through a bunch of capital is a real plus for open source. All right, we're almost here and we're still safe. We are almost to the... we're... to the drop-up, to the edge of the boundary. We can't quite get to the airport. And we're out to the airport. Waymo boundary ends in three minutes. How many miles? I don't know. But yeah, we'll go from a Waymo to an Uber to get to the airport, which I guess that is even more fitting for being in San Francisco. Right. This is how...

We can't quite get to the airport. This is how new technology works. You're so excited, you experience it, you use it, and then when you want to get to the full destination, it doesn't quite take you. You just get a slightly older piece of technology. Yeah. That's fine. Exactly.

What a great way to end this. All right, guys. All right, that's the Waymo Data Council update. We'll have some new stuff coming at you soon. Thanks, everybody. The Data Stack Show is brought to you by Rudder Stack, the warehouse native customer data platform. Rudder Stack is purpose-built to help data teams

turn customer data into competitive advantage. Learn more at ruddersack.com. Rutter Stack is a platform that allows you to create and use your data to create and use your data to create. Rutter Stack is a platform that allows you to create and use your data to create and use your data to create. Rutter Stack is a platform that allows you to create and use your data to create and use your data to create.