Exploring Database Management and the Role of AI with Liquibase’s Pete Pickerill

Hey, everyone. Welcome back here to Tech Drunk tv. You know, my, my next guest reminded me, I think the last time him and I were face-to-face.

Might have been at a enterprise, a DevOps Enterprise Summit does, I believe it was. Yeah. Which That's gotta be maybe before COVID Pete, I think it was, I think maybe 2019 was the last time you And I talked Right directly.

Yeah. Right before COVID, so, yeah. Yeah.

Hey, it's great to, I don't, you know, I'm sorry it's been this long and it's great to have you on here on Text Drunk tv. I hope all's well with you. How are you?

I'm doing great and I'm really happy to be here. You know, my co-founder, Robert Reeves, used to, um, do a lot of the forward facing stuff. They're getting me into more of it now.

Yeah. So we, we will probably be talking more as time marches on Abs. I hope so.

I hope so. I was, as we were talking about off camera, I just saw Robert. I didn't see him.

I did one of these, a Zoom call with him, which I guess I saw him. I just didn't see him in person, as they used to say. Sounds outdated, like recording stuff on film.

Anyway, um, Pete, I, you, you mentioned you're a co-founder of Liquibase. Give people a little bit of your journey and, you know, and the, and we'll segue from that in a Liquibase story. Sure thing.

Yeah. So I was born and raised in Austin, Texas. I've been a part of the tech scene here since the late nineties.

Um, right around 2000, right, right before Roy 2K. Uh, and over the last 25 years, or 26 years, I guess at this point, I've worked at several startups that got acquired by larger companies. There was a logistic startup that got acquired by Neopost in Europe.

There was a security startup that got acquired by Symantec. Um, and then prior to starting Liquibase, there was a, uh, a another DevOps company really, um, as DevOps was kind of taking shape and, and becoming a movement called furnace that managed J two EE CER configuration. That company got acquired by BMC, and it was really our work with BMC that led to sort of the seed of starting a company that does what liquid based does now.

So we were, man, we were helping people manage their JTW server infrastructure a lot more easily with automation, you know, reducing all of the manual errors, allowing it to be repeatable and automated, all the great things that, that you really want from sort of a point solution to DevOps. Um, and as we were doing this, we heard from some of the largest brands in the world, Hey, this is great. This is gonna help us.

This is how it's gonna affect our bottom line. Do you guys have anything planned for the database? And it wasn't just one or two people.

We heard this again and again and again. And as we kinda started probing into those conversations, what we did, what we discovered was while all of these things were being automated around the database, the database itself wasn't being automated. So you had tools like Jenkins, uh, starting that ci cd journey for a lot of companies.

All of the automation around testing, building, assessing, security, producing an artifact, all of that was being written for the application code, no problem. But people weren't doing the same thing for the database. There were no solutions for the database.

It was still a largely manual process. And, you know, not only was it risky because it was a manual process, but it was slowing down the investment. Everybody was making an application code because you can't really ship an application without shipping.

The database changes with it. So, um, we looked out in the world of open source, uh, as part of our due diligence and our initial market validation for starting this new venture. Um, and we discovered Liquid Base, the open source project that next year we'll turn 20.

So it's been around since 2006. Uh, and, and really it exists to help developers organize, repeatably, execute, and automate database migrations in conjunction with application code migrations. And that's what we live to do more and more as time has marched on, it's less about actually automating those things, but providing the, uh, guardrails around that automation.

So making sure that you're, you're maintaining security standards, making sure that you're maintaining audit and observability standards, making sure that the changes that you're introducing to the system are of high quality prior to being merged into the code base. So a a lot of work on the developer front, um, for the developers who are now managing these database changes to really provide the guardrails and the assistance that they need to do this well and make sure that they aren't introducing, you know, performance or security issues, uh, into their database. Um, and that's what we've been doing for the last almost 14 years.

The company will turn 14 next year, next to April. Um, so yeah, I mean, that, that's, that's really why we exist. We want to help you manage your database changes with the same level of visibility, security, safety, and, and, um, efficiency as every other, uh, component of the application stack.

Absolutely. com in, uh, 20 13 20 14, we first published early 2014. Um, of course, liquid based was known as data at the time, right?

Yeah. Liquid Base was the open source project. But really, you guys kind of pioneered DevOps for databases, because you're right, databases were kind of the redheaded stepchild of, of the whole IT stack, right?

You know, the, I don't know why, well, I do know why. 'cause I've seen it over the last three or four years where data has become pri primacy again, right? Data is, yes.

For a little while there, early two thousands all the way through, let's say 2020, we seemed to have laws. It felt like the Bill Clinton campaign, remember, it was keep, it's about the economy stupid. We remembered that sometime around 2020.

It was about the data stupid, it's about the data. We could put all of this effort into our application and our application processes and and development, but really it was about the data and, and the databases, of course, the container for our datas. And, and so there's been a refocus on database and the data, the primacy of data.

Yeah. So yeah, liquid base is wr ridden that kind of wave now, right. Or rode that wave.

Yes. I guess is the right word. Yeah.

Yeah, absolutely. So, so that's, that's kind of what, what, you know, over the course of the company we've noticed as well. And I think that 2020 date is really kind of an important one because I think that's when we really started to see the initial, um, started to get a hazy picture of what AI was gonna do for us in the future.

And obviously you can't really train these l lms, you can't have, um, robust, um, you know, robust LLMs that don't hallucinate as much without high quality data and understanding the lineage of that data. It's always As good as your data. Yeah, Yeah, absolutely.

So, I mean, I think what we saw is, initially people were trying to fix a functional problem, but now they're trying to fix a strategic problem. They're trying to stay competitive in a world where AI features and AI tools or launch popping up like crazy. Um, and, and you know, there's a real sort of, I think there's, there's a pretty, um, uh, uh, sharp increase in the cadence with which people feel like they have to move in order to stay competitive.

So, um, you know what, we found our customers that, you know, have been customers for several years and kind of have the operation of managing databases on rails in their automation already. It was really a lot easier for them to start, uh, you know, launching these AI initiatives because their data was already, you know, there was a high level integrity because it was being consistently managed across the pipeline. Um, and, and those were the things that kind of are, are driving our customers now.

It's not so much about the initial promise of DevOps, but it's about supporting these, these AI initiatives that are really just kind of have spiked the demand for high quality, um, understandable data. And, and that's, those are kind of the problems that we're solving with our customers now. Love it.

I love it. com is the best place mm-hmm. To jump in and, and start learning about Liquid base, familiarize your yourself with Liquid Base or see what's new in Liquid Base, if it's been a while since you've checked it out.

Um, we've got, from there you can launch to, you know, our GitHub repo. We can get to our documentation. We have a whole site on contribution and, and things like that.

We have forums where, uh, that are pretty active, where our users go to discuss usage of liquid base and troubleshoot issues. com is your gateway to all of those things. So I, I would definitely start there if you wanna learn more about Liquid Base.

Excellent. All right, Pete, I'm gonna shift gears a little on here. I wanted to talk about, well, it's actually what Liase calls the database delivery gap, fixing the last mile of DevOps.

And a lot of this comes out of a, uh, uh, report insight report. You guys did the 2025 state of database DevOps. Um, actually before we even just start discussing, for anybody who wants the report, they can get it, I'm assuming, from the liquid based site.

Do you have like a friendly URL to give them or just go to the front page? I, it, it's on our blog. So I go to liquid based slash blog and just look for state of database DevOps.

com/beal/state-of-database, dash DevOps dash report as of dash one H dash 2025. But it's probably easier just to go to the site In the blog and look up state of report. Exactly.

Alrighty. Um, spoken like a real database person there feed. Uh, so talk to us, what, what's in the report here about the database delivery gap?

Yeah, so really we, we kind of touched on it earlier. You know, the more things change, the more things stay the same. There is, uh, because there are, you know, factors like AI and really just, you know, some, some of the, uh, the later companies in, in the process of adopting and, and building out their DevOps practices.

Um, you know, it's really driving this focus on the database because there are a lot of, of, of proven tools and patterns and things like that for everything. But the database changes, but kind of getting the database changes to move and lockstep with those database changes and to benefit from the same automation is really what the biggest gap is for a lot of our customers. We'll see folks who have homegrown solutions, or they're using an ORMA framework of some kind, but frameworks don't really provide all of the, um, necessary governance and observability benefits that you, um, that you really need in this day and age of ensuring that your data's in, in high of high integrity and that you're satisfying audit and compliance, um, needs.

But it's, it's really what we're seeing is that, um, you know, not only is there a gap in process, there's also, uh, kind of a gap in, in understanding and training. So, um, yeah, just to kind of cover some of the highlights of that report, obviously we were talking about this a second ago. AI and ML are really driving sort of, you know, an explosive, um, need for data, a demand for data that unlike we've ever seen before, we need more of it.

We need it faster, we need it, uh, to be of high integrity, and we need to, um, make sure that we are managing the process of moving it from our transactional databases to our data lakes or our S3 buckets. Wherever, wherever we're analyzing our data, we need to make sure that process is on rails. So AI is really kind of driving a lot of that adoption.

And this was, this report came out of a survey. We surveyed, you know, a lot of people, customers and, um, OSS users alike. Uh, and, and in that we found that 78% of those respondents said the biggest challenge they're facing right now is rising to the challenge presented by AI and ml.

So, um, but what we also noticed is that the teams who've been doing DevOps for a while and kind of fell into our mature cohort, um, it was really just kind of business as usual. They already had their databases on the rails. They had all of the audit and lineage information they needed.

They were sure that their processes were fluid, there were guard rails to keep them from breaking. So really it's just like, okay, we've got more data that we need to ingest to move from A to B once they have database DevOps sort of, you know, baked into their current processes. So that's a lot of what we're dealing with right now is com.

Those, those companies in the 78% that are trying to get there more quickly, they're having a real tough time with that. But with this new proliferation of, of tools and, and, you know, processes and different data platforms and different patterns for managing that data, um, you know, we also found that a lot of the practitioners feel like they, they're not really getting the training. They, they need.

They're being given the tools they need. They're being given sort of the initiatives and the, the, the milestones they need to hit. But what they're not getting is, is training or, or advice or, or really just kind of that first step in DevOps, which is, you know, let's understand our entire process end to end so we can all bring our different perspectives together to build the best possible process in terms of speed, safety, and reliability.

So, um, yeah, I mean, it, it, it, it's really just a lot about, you know, we're seeing a lot of new tools, we're seeing a lot of new patterns, we're seeing a lot of great innovation in the space, but I think that, you know, what we're hearing from, um, survey respondents what we cover in the report, a lot of people feel like they aren't getting sort of, we aren't investing enough in the human aspect of that, the humans that use the tools. So I thought that was really interesting as well. And that's something that I've been talking to people about, um, you know, pretty consistently whenever I get the chance to speak to them, for sure.

I love it. Pete. Sounds like a great, a great report.

We said where, where it's, uh, available. What else going on at Liquid Base? You guys gonna be at any shows anytime soon or maybe CubeCon or something like that?

Yeah, I think, I think we'll boost on the ground at, at CubeCon we're looking at, uh, reinvent also, and we're, we're also mm-hmm. Um, I moved into more of a developer relations role this year. We're focusing on local events as well, so various dev Austin's a hot area, PCDs or like that.

So yeah, I mean, I think, we'll, we'll be around, we'll be pretty visible, um, at, at the major meetups. And then we're doing some regional stuff too. So I would love the opportunity to talk to everybody, anybody that wants to talk databases or DevOps or really anything, um, and I'm pretty accessible.

com. I make it easy. I love it.

Pete, it's been a while. Don't, let's not wait five, six years for you to come back on Tech Trunk tv. Okay.

Agreed. Agreed. Yeah, it was great.

Great talking to you again. All right. I look forward to the next one, Alrightyy, Pete Pickerel, co-founder Liquibase here on Techstrong tv.

We're gonna take a break. We'll be right back.