Why bad data didn’t matter until now

David Sweenor 0:00 Hello everyone, and welcome to the data faces podcast on location. I’m David Sweenor, founder of TinyTechGuides, and your host for today’s show. In this show, I talk with people who are making data analytics and AI work in the real world. What’s exciting, what’s messy, and what’s coming next. Today I’m here at click Connect 2026 with Brendan Grady, Executive Vice President of all things AI and analytics at Qlik. He runs the business unit, and we are going to talk about AI today and all things click looking forward to it. Welcome to the daily faces podcast. Thanks for joining.

Brendan Grady 0:34 Thanks. Thanks for having me excited to be here.

David Sweenor 0:37 So can you just tell us a little bit about yourself and click for people who may be living under a rock. Speaker 1 0:43 So obviously, got my name Brandon Grady. I run the analytics and AI business unit. I have responsibility for product management, product design, R and D go to market strategy and everything that is involving that side of

David Sweenor 0:57 our business. Okay, that’s a tall order. Yeah, it is all right. Well, part of this, the name of the show, is databases. So to get behind, you know, and uncover who you were, maybe before you were into data. So I got a little icebreaker for you. What was your first job?

Brendan Grady 1:14 My first job prior

David Sweenor 1:17 to, prior to the LinkedIn profile existing. Speaker 1 1:22 I’m gonna give you my first if that’s okay, but I’m gonna be my most recent Prior to there. Is that? Okay? So my most recent job, prior to me getting a real world job, was, I was a tour guide. Okay? I gave The Sound of Music tour in Salzburg, Austria, well over 300 times. So if there’s any guy you want to know about the sound of music. If you like it, don’t ask me, because I will ruin it for you. But it was, yeah, that was my job. It was fun, you know,

David Sweenor 1:48 fun fact, I’m living in Burlington, Vermont right now, and the van Trapp Lodge is in Stowe Vermont. Speaker 1 1:55 It is the von Trapps. It’s up there. Yeah, it’s, I have to say, I’ve only been there once, and just didn’t really need to go back, having been exposed to the movie so many times.

David Sweenor 2:04 All right? So you know, you and I have a long career in this space, and you can tell by my lack of hair, but you know, you’re on the GTM side, and I went to the dark side, from being a practitioner of analytics and data warehousing to the dark side of Product Marketing. So you know, back in that time, sort of, what are the lessons learned, what worked and what didn’t? Speaker 1 2:26 I always start with this, everything, everything old is new again. It just feels like it’s, feels like we’re back in when you and I probably first met, way long ago. Everything was big data, right? Big data, big data. By the way, the data is still big, okay? It’s still there. So, so if I think about what we what we got right back then, I mean, ultimately, was all about people need to make decisions with data. That’s just the reality of the situation. And companies that we worked at together actually were at the forefront of that, trying to bring data, making sure they had the right data, the right data quality. They did some major acquisitions on some of the big bi players back in the early 2000s and realizing that 25 years ago, and it’s hard for me to say that, right? So almost 25 years ago, they knew that data and analytics was going to be important for where the world is today. Okay, it was that foresight. I think some of these companies had to really, I don’t want to say predict, but really have an idea that this data thing is not going away, and it’s going to be critical to everybody moving forward, and is even more critical today than it was 15 or 20

David Sweenor 3:32 years ago. All right, so speaking of data, everybody talks about data quality and getting it right, like, why hasn’t this been fixed? Why hasn’t it like, we’ve had data on clay tablets and Papyrus? It’s still a mess. From what I can tell, it definitely is still Speaker 1 3:51 a mess. And I mean, I think it gets down to a couple things. It’s hard, it’s hard to fix. That’s one, two consequence management. If you have a number that appears in a spreadsheet, for example, or some little visualization and it’s off, you just go fix it, right? So that’s that’s fine, but in today’s world where there may be an agent running around using said data and getting it wrong, the consequences of getting it wrong are going to be catastrophic. So the reason why people may not have fixed it was in a manual world. Then it really mattered. Didn’t really matter. You can recover, right? That’s but even then it was still hard, right? Because is it, is it it’s job? Is it the business’s job? Is it both, or is it nobody’s job? So that’s think there’s a little bit of organization technology, and then there were no consequences for getting it wrong,

David Sweenor 4:46 if you could even detect that there’s an error, oh, God, spreadsheet, you know. So, you know, that’s the one of the things that strikes me. So in this, this world of agentic AI and AI agents running about, you know, you might be like Sween. Six or seven or 10 steps down the line before you realize you messed up your your business? How to organizations reconcile with that? Speaker 1 5:08 Well, yeah, and one of the things that we’re really focusing on at click is to make sure you don’t get into that trouble, right? So I think clicks legacy in where we started as a desktop data discovery tool with this analytics engine, right? That helps you find patterns in the data that you can’t normally find. It’s just not possible with other tools. So that’s one thing we’ve invested heavily in the data pipelines and making sure they’re resilient, so that you know you can trust them. And then we’ve also really go down the path around this trust score, for AI, really making sure that it’s quantifiable, so that when you get into this agentic experience that you’re not, I don’t know how the polite way to say this. You’re not giving knives to three year olds, right? Or running with knives on a slippery floor, sure, but we’re really trying to make sure that you can quantify. Yes, this data is, in fact, valid. Yes, it is fresh, and because of our engine, we are going to find patterns in it and references that you wouldn’t normally find in a typical analytics tool.

David Sweenor 6:08 Okay? And you know, you talk to clients all over the world in your role, and where would you say the state of adoption, agentic adoption is, you know, I see these early on demos of agents. Hey, it can order your groceries. I’m like, or schedule your calendar. I’m like, That’s not that interesting to me. So from an enterprise level, kind of where, where are we in the maturity curve Speaker 1 6:35 for what’s prior to stage zero? I mean, it really, there are customers that are trying things out there, right, surely and but oddly enough, you brought it up just a few seconds ago. Data quality is preventing them from getting there. So we’re seeing customers go after very, very low risk use cases. And it’s not ordering coffee for the office, but there are really low risk use cases that they’re trying to go after from a large scale production. We’re so early days. And if you think about how long it’s been since agentic really started taking off, the internet, took 10 years, right? 20 years, 30 years to get going, right? We’re a year in, right, okay, a year and a half in. So I think we’re so early days get customers with data problems, so not necessarily going to take off as quickly as some people thought it

David Sweenor 7:26 was going to. So Brandon, when customers think about llms, you know, there’s a number of them out there. How do they future proof themselves? Say six months ago it was open AI today, it’s all Claude. Claude. Claude tomorrow, it’s going to be something else. So how should organizations think about this? You want to Speaker 1 7:46 so you want to be working with a vendor that can actually approach this from an open standards perspective. So we at Qlik, we really focused in on supporting every LLM, right? We want to make sure that if customers want to use chat GPT, they can use chat. Should be T, they want to use Gemini. They can use Gemini. We’re seeing strength in Claude. I think you and I were discussing that, right? Claude is pretty strong, so we just make sure you’re working with a vendor that’s open to using other, other llms out there, not just forcing you to

David Sweenor 8:17 use one. Okay, that makes a lot of sense. So let’s talk about the, I think, was formerly known as the associative engine, the analytics engine at Qlik. Speaker 1 8:24 What’s that all about? Oh, the associative engine was what I didn’t know at the time when I joined the company seven years ago. It was the reason I joined the company, really, I was competing. I was competing with another analytics at another analytics vendor, competing with click and losing constantly. And so when I got here, I dug into this thing that was at the time, was called the associative engine, which we’ve renamed

David Sweenor 8:49 to the analytics engine. Speaker 1 8:50 Analytics engine got it And fundamentally, what it does is, instead of creating a SQL query and going to get you just the data you want to answer some specific question, it goes and gets all the data and finds associations and interconnectivity and interrelationships between that data that you didn’t know. You didn’t know. That’s the most amazing thing, okay, it’ll help you identify problems and relationships that you didn’t necessarily know existed in

David Sweenor 9:19 your data, the unknown unknowns, the unknown Speaker 1 9:21 on, yeah, we go way back on, right? Yeah, the unknown unknowns. It’s a little bit of that, and it surfaces it in a way that it’s works like a human brain. So just like we ask questions and we go down these certain paths, it’s designed to allow your brain to go down a path to get to an answer or make a decision. In the agentic world, you think about the power of that, it will allow an agent to go down different paths trying to make the right decision with relationships that may not be recognized by other tools. You know, I love

David Sweenor 9:53 that, because you know back when, way back in the way back machine doing dashboards and. AI if unit builder or called somebody to build you a dashboard which had a predefined path, it didn’t work for you. And what I see, and I’ve worked for a number of organizations, they have 1000s dashboards, not a lot of insight. So this sounds like it’s really helps you, whatever question I have, it helps you find that. It helps. That’s the known ones and the unknowns. Like, oh, maybe you’d be interested in this sort of, like a Netflix recommendation. I know it’s not a recommender engine, but kind of like that. Speaker 1 10:29 It is kind of like that. It’s surface, and we actually do it visually, right? So that’s really interesting. Thing is, today we do a visually so it’ll show up as this green, white, gray color. And so the gray sort of things you need to be thinking about. Let’s go kick the tires on that in the agentic world, we’re serving this up to help agents understand that there’s a relationship here that you need to go explore before you take action. Okay, that is extremely powerful. And from

David Sweenor 10:53 the data perspective, you know, I know a lot of people were still talking about rows and columns. There’s this whole other world of PDFs, mostly PDFs, Audio, Music, unstructured data are does that become more important now with llms, it seems like they eat that more than they do this sort of the rows and columns, the traditional data that might have been stored in Speaker 1 11:21 a data warehouse, yeah, where the llms are strong, obviously, is in that in that unstructured data, right? Because they can bring it all together. Where they fall down pretty hard, actually, is on that structured data. In today’s world, whether it’s structured or unstructured, and we’ve been at this long enough to know that there was a world of unstructured and structured. It doesn’t matter anymore, right? People just say, I need to get something done and I want to ask a question. They’re not saying, Well, let me go to my structured data tool and go do that. Right? Let me go to my unstructured data tool and do that. They just want to go into their interface of choice, ask the question, get an answer, take an action, and it better darn well be right, because if it’s not, there’s a massive cost of having a

David Sweenor 12:04 mistake like that. How do how do you sort of think about that? I’ll tell you a little story. So I was helping a client with a website, and I don’t know anything about fonts. There’s plenty of designers out there that have big books of fonts, but I said, Go tell me why the website looks weird. I think it’s the font. I had a hunch. And I said, go look at website, A, B, C, D, E, and it comes back with some analysis. It looked great. It dutifully did it. And then I said, Did you really go look at the websites? Oh, I’m sorry. Let me. Let me go. Do it again, and it came back slightly different answer. Then I went and looked with my eyeballs, and it still didn’t do it. These, some of these things have a like a a mind of their own, even though you explicitly try to fact check, tell it not to hallucinate. So I guess my broader question is, hallucinations are a big issue in this space, right? Speaker 1 12:59 Yeah, so, and the way we’re thinking about it is, where we’ve really invested is around intent detection, right? So what is the question you’re really trying to ask? Right? Because, as you just sort of pointed out, you’re sort of going down this path where you’re trying to get one question answered, but your intent is really something else to get done. So we’ve we focused pretty heavily on that. That’s one thing. The second thing, it comes back down to our engine. Because of the way our engine was built, we can actually surface up structured data inside an LLM and have it be 100% accurate, right? That’s the really interesting thing. If you’ve ever tried to dump a spreadsheet in and say, Hey, Gemini Claude, chat, GPT, go calculate something off the

David Sweenor 13:40 spreadsheet? No, it’s not a calculator. Speaker 1 13:43 Oh, not at all. And it’s but it’s really pretty, right? The answer is, amazing. Great. Looks great. Totally BS, it’s all it’s all crap, right? That’s right. And the next thing you know, you’re showing up to the board with all incorrect numbers. So where we are really focusing on is getting the intent of the questions that you’re trying to get in the LLM, and then servicing your data from your click application, so that you know that it can be trusted. And then, even more importantly, we apply this trust score to that data, so that you can quantifiably say yes, it was 90% accurate, 80% accurate, or whatever it may be, yeah, that’s sort

David Sweenor 14:17 of Lou I love this idea of a trust score, because in the past, people look at maybe some visualization of data dashboard, what have you? Just take it as dogma. There was no confidence or whatever. I’m like, Oh, well, it’s on the dashboard. It should Speaker 1 14:33 be, right? Yeah. I mean, it’s the trust core for AI that we announced a couple years ago. And I was really focused on making sure that people don’t get into trouble with that. And I always encourage anybody that’s listening to this to just go have a little fun. Go, go Google bi or AI failures. Yeah, it’s super fun, like it’s, I mean, it’s so scary, really scary, if you’re in this line of work. But what do you what some of the stories you get? I mean, you go back to like, the United Airlines, right? Or United took a massive hit in their market cap because their data around sentiment was wrong and right. And so you look at that, you look at Fannie Mae, this is 20 years ago, but there was one field in an Excel spreadsheet that took the market down. So that’s the consequences of this stuff. So really focusing in on getting a quantifiable measure. Say, look, you can make a decision off of this, and you can have confidence that the data is right. We feel that’s extremely critical in this world moving forward. Yeah, so,

David Sweenor 15:30 so important. So, you know, we talked a little bit about dashboards. I know you have this concept called decision intelligence, and some, some of the industry analysts we know have been talking about this for quite a while. Is it real now?

Brendan Grady 15:49 Dashboards are dead.

David Sweenor 15:51 Dashboards are dead. We have a scoop. You’ve heard it here Long live dashboards Speaker 1 15:56 with that. It is, it is fast. I mean, look. I mean, if you’re listening to us right now, you may have picked up. We have been in this space a long time, and the reality about dashboards are this, it’s really simple, right? Is that dashboards should be used for making decisions. Dashboards in the way that we’ve grown up with them as a destination is probably going to go away, right? It’s not going to be a place you go to, but because, fundamentally, humans learn in one of three ways. They’re visual learners, right? They’re auditory learners, or they’re kinesthetic learners, dashboards will be surfaced in a different way. So you may get a visualization as part of your LLM that comes back, that’ll be great for some of the visual learners, right? Other people want to say, don’t show that to me. I don’t want to see a visualization. Just tell me what to do with the data, right? So I think the concept of the data that’s been in this dashboard was that there’s these nuggets of gold and platinum that you can use that will continue to exist, but just the mechanism you get that information is probably going to be starting in an LLM or an agentic

David Sweenor 16:59 tool, right? That’s so it’s always been sort of their Nirvana, yeah? Like, I get some insight, then you had to go figure out what to do. Now you’re getting that prescriptive recommendation. Maybe you should go turn this knob and do this and watch your business. By the way, I got something to help you track the performance of your business. Is that the idea? Speaker 1 17:18 That is the idea and so what? So at click, we are customer zero. We use our own products. All right, that’s good. We always use our own products. And so the way I’m using our products, I either use one of our capabilities, which is bringing structured and unstructured together, our agentic experience, or I tend to use I go to say it, I go, I use Claude. And so I go in and I ask questions of Claude, what I would normally do in the old bi world is give me information about my business performance. Yep, today’s world using sort of this intent detection I talked about and and some of our MCP capabilities, they say, tell me about my business and what you think I should do right radically different questions. And that’s the world of people who grew up going to a dashboard. That’s the world of changing.

David Sweenor 18:05 Yeah, I see it now in my everyday life. I do have, like, maybe this is a philosophical question, but how much of our brains are we outsourcing? Oh, boy, to these, to these llms. Do you think there’s a risk? And I always, I’ve asked a few guests this, and there’s, there’s two dimensions. One is, like, your personal self, and I have young kids, and like, Well, why do I need to learn anything? You know? Probably, like, maybe it parallels when the calculator came out, or word processor or whatever. But I’d love your perspective on sort of brain outsourcing in the future of the Speaker 1 18:43 young ones. It’s scary. It is. I do think there’s a risk. There’s a risk of of outsourcing. I’ll call it critical thought, right, right to say, well, I asked the LLM, and it gave me an answer. You know, I’ve asked the LLM a bunch of questions before. It gave me an answer that I could have come up with if I with if I had applied my own brain. There definitely is a risk. There’s some studies, you probably saw, the MIT study, where there’s, they’re talking about the brain atrophy that happens, if you follow so Sol Rashidi at all up on on LinkedIn, she’s been talking about this a little bit, right? Yeah, I think the philosophical conversation I have kids as well. Got a 24 year old and soon to be, 20 year old? Sure? Yeah, I’m watching what they do with these llms, and I’m watching a little bit of critical thought go out. But I do see there is hope, right? There is hope, because what I’m finding is that people are starting to use these llms to augment their thinking, versus do the thinking Right, right? I use it for a lot of validation. I’m seeing some of our customers using it for validation versus go create me something. It’ll be interesting. Come back in a year from now and see what actually happens. Because, as you know, there’ll be between now and the end of the year, it’d be like 300 more changes exactly.

David Sweenor 19:59 We moved on for. Claude probably, yeah, Speaker 1 20:01 it’ll be the next model, right? I mean, that’s it so, or be pretty fascinating, yeah,

David Sweenor 20:05 I find there’s like, such a newness with it. You want to, I’m gonna use it for this, this and this. And there was another MIT study that, you know, was like, hey, you know, below average performers would come up to sort of mediocre performers, but high performers got better because they didn’t outsource all of their brain. They knew when to use it and when not to use it. We have 10x Speaker 1 20:25 developers here internally. They it took real they took good developers, and it just skyrocketed them. We’re starting to see the people that were already doing really well. You’re seeing more acceleration at the top end. That’s what that’s been our experience. And it’d be interesting to talk to anybody who has a different experience. Is it bringing sort of average and Lower, lower performing employees and making them good great? Who knows? We’ll see our experience has been 10x developers.

David Sweenor 20:53 Yeah, I always, always tell people that you know, don’t live in the AI bubble like you and I, I’m like, AI is great if you know what you’re doing, yes, you don’t know what you’re doing, you’re going to get into a lot of trouble quite quickly. Like, if I asked it to write something about data analytics, I’ll know what’s right, what’s wrong. Put this in, take that out. If I ask it to write something about art, I don’t know anything about art, so I would have to take it at face value and give it to someone that knows about art, and they’d be like, this is total crap. Because I mean, you and I’ve Speaker 1 21:20 spoken to data scientists over our career, too, right? I mean, sort of giving mere mortals access to predictive analytics 10 or 15 years ago was super scary, right? It was really scary. Is it as scary today? Or, I don’t know, it’s a maybe

David Sweenor 21:36 the citizen data scientist will come back, and Speaker 1 21:38 maybe the citizen data scientists will come back. It’s Wow. I can’t why you went there. I did. That’s crazy.

David Sweenor 21:45 So I know click has a MCP server. Can you tell us a little bit about that and why you decided to go that route. Sort of opens it up to other parties. I think it does. Speaker 1 21:57 And look, I think the we’re in a world right now where standalone vendors are going to have a hard time of anything, right? People that don’t, that don’t see them as part themselves, as part of another broader ecosystem, because there’s never going to be One Ring to rule them all. Despite what some of these data platform vendors will tell you, the hyperscalers, everything’s going to be on arts. No, it’s not. It’s always been like that. It’s always been like that. It’s just, it’s it. I think it’s just becoming a little bit more critical at this point. So we went down the MCP path for two reasons. One, to be open, right? We wanted, we wanted to be able to give people access to our analytics engine, which has been a game changer for Qlik for many years, plus give people access to the trust, the trusted data foundation that we put in place. But the really interesting thing about that is, I believe, personally, and this is a strong belief of our design team, my product management team, people are going to want to work within the tools they’re comfortable with. And that may be Gemini, it may be chat, GPT, it may be Claude, it may be something completely different, but the reality is, the best way for us to do that was open, open. It Up the MC pay. And we’re seeing right now, we’re seeing about a 5050, usage between our own interface and MCP on our new agenda capabilities. Pretty interesting.

David Sweenor 23:18 Okay, okay, that’s great, yeah, because there will always be people that, to your point, they want to work a certain way. I’ve tried training marketers how to use Claude co work. Or I’m like, Hey, Claude code is even more powerful about that. They’re just so intimidated by that sort of interface. You know, co work is, is more more suitable for them. So, you know, we are here on location at Qlik connect. And what are the key themes we’re gonna we’re gonna see and hear about as we move forward. You know, throughout the year, from from Qlik trust, it’s gonna be one of them. Speaker 1 23:54 Context is gonna be another one, and flexibility is gonna be a third. Those are three things that you’ll hear a lot about, because ultimately, can’t have context that you can’t trust what you’re looking at, right? And we also want to make sure that our customers understand that we’re not trying to lock you in, right? We continue to be very flexible, right? Think we’re demonstrating that by our MCP announcement sure just came out, that we want to be open. So those are, those are three key, key themes. We’re obviously doubling down on the agentic experience. Yep, right. We know that agentic is going to fundamentally change the way we data professionals have worked, and it’s going to change the way every human works. So you’ll hear that as a key thing, and then you’ll also really hear, when you get into the capabilities level, on the data side, heavy emphasis on trusted AI ready data, that’ll be a big theme for us for the rest of the years. And a lot of our agentic experience will be around making that those data pipelines even stronger than they are today, and making, I don’t like the word agentify, but I’m going to use it. Why not? Why not?

David Sweenor 24:59 Right? We’re around the pipe. God, first agentify, gentify. There we go. So you’re

Brendan Grady 25:02 gonna, you’re gonna see that a lot on our data

David Sweenor 25:04 pipeline side as well. Okay, and, you know, from when I knew click, you know, was primarily a sort of a BI tool, your portfolio is significantly broader now, and it’s, I think you have the the tools to make this a reality Speaker 1 25:21 that, yeah. I mean, if you look at, if you look at click, I mean our history, we are a, we are a company that’s founded in Sweden in 1993 yep, that’s insane to me. The founders had the foresight to put this engine in place. Sure, and that engine is something that’s going to be critical for context in order for the agents to make the right types of decisions and not get them wrong. So we we’ve built on top of that. In addition, obviously, we brought a couple companies together over the years where we get all the things that go around, data pipelines, all the things that go around Data, Data Trust. We bought an automation platform a few years ago so that we can actually take you from absolutely raw data to acting on that data, this reality that we’ve been speaking about for years, and decision intelligence and decision and Decision Management, we’re here, and we can do it, and we can do it today. That’s a really exciting thing.

David Sweenor 26:12 That’s amazing. So maybe just some parting words of advice. So for for people who are there’s just a lot out there. Yeah, we talked about data quality, we talked about MCP servers, we talked about agents. How does someone make sense of all this? And kind of, what’s your advice on how they should think about it? And potentially, you know, what would be their for your recommendations to avoid some of the mistakes we’ve all made in the past? And, you know, put their best foot forward. I would say, Speaker 1 26:43 embrace, embrace these new technologies is the first thing I’d say, it’s scary. I had a customer come to me after we demoed the ability to build an analytics application through Claude in 30 seconds. Okay, and this person has built his entire career on writing code to build analytic applications in multiple products. And the question he asked me is, he said, Am I out of a job? My answer is, No, you’re not out of a job, but your job will evolve. It’s changing, changing. So my advice as a data professional, you know the data right? That’s where you grew up. You know that data probably better than some of your business users embrace that. Think about becoming that data product owner, becoming an expert in that data, and being that trusted guide as everybody’s going down the subject experience. That’s the real opportunity for data professionals.

David Sweenor 27:33 Okay, yeah, I think there is a lot of opportunity there. And I know that said that was my last question, but I lied. Oh, boy. Just another thing I was thinking about is, does this new world expand the aperture for people who can participate in data and analytics practitioner, I know like they’re not going to be experts, but maybe they can do so. I’ve seen marketing campaign owners who typically would rely on an analytics team say, oh, I can do some basic preliminary analysis now myself without relying on someone else. So I’m just curious on your the aperture, does it? Does it open it up to a broader set of people Speaker 1 28:15 or and I’m glad you didn’t say the word democratizing analytics? No. I mean, we could.

David Sweenor 28:19 We could democratize analytics with a citizen dataset. Speaker 1 28:22 Exactly, we should create a time machine and go back way back, for sure, it absolutely is opening it up, but it’s opening up for two for two reasons. One, that the technology is there, right? The ability to make this easy, we are finally at that point where anybody can go in and ask a question of this and get the answers they want. That’s the really interesting thing, and that you can trust it. The other thing is, the expectations of everybody who’s in the workforce is fundamentally shifting. I mentioned I have two daughters, 119’s and 124’s sure they don’t want to go to a dashboard, but they they have to get answers to the questions and their expectations, whether we like it or not, has been changed because of chat GPT just a few years ago, everyone expects that you can get any answer from that interface right, and the technology is catching up to be able to do that. Okay? Well, I

David Sweenor 29:14 love that. Brendan Grady, Executive Vice President of all things AI and analytics at click, click, connect streaming online. If you can’t make it in person, encourage everybody to check it out. So Brent has been a great discussion. So I appreciate you being on the databases podcast on location. Thank you very much. Good to

Brendan Grady 29:33 see you again tonight. Cheers.