100. Making Agentic AI Real for Business Intelligence with Qlik – Tech Field Day Podcast Spotlight

Generative AI is more connected to the world of business intelligence and data analytics than any other area of enterprise IT technology. This episode of the Tech Field Day podcast, recorded prior to Qlik Connect twenty twenty-six, features Brendan Grady of Qlik joining Brian Budden and Gina Rosenthal to discuss how this market is advancing. Welcome to the Tech Field Day podcast, where we bring together a group of IT technical experts to discuss a single idea about key concepts in the industry.

This podcast features a variety of perspectives from members of the Tech Field Day community and is often recorded in association with one of our events. Tech Field Day is part of the Futurum Group, and this podcast is also published on our sister company's site, Techstrong TV. On this episode, as we look forward to Qlik Connect in April, we're talking about the ways in which business intelligence and data analytics have changed in the generative AI era.

But before that discussion, let's meet who's on the panel today. Hi, my name is Brendan Grady. I am the general manager of the AI and analytics business unit here at Qlik.

Hey, my name's Brian Budden. I'm the managing director at BB Data Services, and I'm a nine-time Qlik Luminary and partner ambassador. Hey there.

I'm Gina Rosenthal, and I am the founder of Digital Sunshine Solutions, where we help businesses make sense of their very intense, technical products with AI and cloud and that kind of thing. Looking forward to Qlik again, so ready for this conversation. And I'm Stephen Foskett, organizer of the Tech Field Day events here, and a Qlik Connect veteran a bit, I would say.

This is gonna be my third time joining the crew at Qlik Connect, and I'm really looking forward to coming back, to seeing, Gina, Brendan, Brian again. Now, for those of you who are listening and maybe aren't, you know, aware, so Qlik is one of the leading companies in the, data analytics, business intelligence world. Um, they have a really kind of a fanatical, fan base in this world, and it shows at Qlik Connect.

The, the people that show up, the kind of discussions we have, the kind of goofy, fun stuff that happens behind the scenes, it really is, different from what you might expect at some sort of dry tech conference. And Qlik Connect also has some other interesting and, and fun elements. Um, you know, the show floor is really, exceptional, and maybe we'll talk about that a little bit.

But, one of the things for me that, that gets me the most is, like many conferences, there's an opportunity to talk to people and learn from them. That's really why I go to these events. But Qlik really facilitates that, both in terms of the folks who work for Qlik as well as the folks who are attending the event.

And I've had just some absolutely incredible conversations, you know, here and there and around the corner with, you know, attendees, whether they are presenters or staff or, frankly, even better, some of the folks who are coming and bringing their expertise and their learning from, what they've done out there in industry. Of course, last year we spent a lot of time talking about generative AI, and this year, we expect that to be even more of a, of a focus area, a focus for Qlik. So, Brendan, let's start with you.

Uh, you've seen the in- you know, the, the secrets. You've seen the, the schedule, the agenda. You know what's gonna be presented, and you can't talk about it.

But I bet what you can talk about is some of the general the- themes. So, talk to us a little bit about this. What can we expect from Qlik Connect, and what does that mean to people out there that are dealing with business intelligence and data?

You're right, I can't really talk a lot about it, so this could be a very quiet section, Stephen. However, I am gonna give you a little bit, a little bit of a sneak peek of what we're going on here. All of us that are, that have been in this domain for many years have witnessed what has sort of happened with people's user experiences, and it all started with OpenAI's ChatGPT being released a couple years ago now, right?

It, it feels like a, a lifetime ago. So what you can expect out of this particular event is to hear about how people's user experiences, whether in, they're in the data engineer domain or whether in the analytics domain, those experiences are going to change. But that, how Qlik is in a position to support that evolution to this new world, which is really all about agentic AI.

And then Brian and I were just talking about how every year there seems to be a shift in this technology, and I think, Brian, you and I agree that it's sort of every couple of months or every month there seems to be a massive shift. So there's some things that we did since la- that have happened since last year that we'll really be able to hone in on, specifically around agentic experiences and what that means for all of our customers and people that work with Qlik. Yeah, Brendan, as I, I think that's relevant.

And as you talked about before, I think last year at Qlik Connect, there was a lot of groundwork being laid for AI, right? There was a lot of, there was a lot of theory being passed around, and there was a lot of understanding about, well, how can we prepare for the onset of AI? And now, now AI has really migrated past that whole ChatGPT section that we talked about there, where people were just hacking in stuff into their ChatGPT windows, into something that is really authentic in terms of AI and BI.

So what, what are some of the big steps that you've seen in that interim period? Because like you say, this market is moving incredibly quickly. So what, what are your big observations in that area just now?

So the big observations are, are still some numbers that are out there, Brian, around the em- or percentage of organizations that have actually successfully implemented agentic AI. Um, no big secret for everybody listening to you, Stephen, it's not a very high percentage, right? And that's one of the things.

So where we've really been focusing on is let's get out of the science project area and sort of the laboratory and take advantage of what Qlik has always done extremely well, which is really make sure that we can give some context around what yourTrying to understand around your business, 'cause when there are humans involved, it's one thing. We can look at it and determine when the sygentic world starts kicking in, there may be autonomous agents making decisions on your behalf, and if you don't have the context and you don't have data that you can trust, those agents are gonna make really bad decisions on your behalf. And then you know how, you know how challenging that could be.

And then Gina, I know sort of your space, you trying to make it real, so it'd be really sort of interesting to get your perspective on sort of how we're thinking about it and some of the things that Bryan and I just brought up there. No, that, I'm looking forward to... That's one of the reasons I liked going to Qlik.

W- in addition to what Steven said, just, there's so many... It's very obvious you put a huge amount of emphasis on your users and making that user experience amazing, which is great. Um, yeah, there's so much hype, it's ridiculous.

Like, how do you know what to do and how to do, and yet everybody's doing it. And we talk about this on our podcast a lot, there's, even though there's so many horror stories, there's so many good stories. That's another thing I enjoy coming to Qlik about, people finding their way through the, the hype and honing in on their data and their people and their processes to make sure they can figure out what these agents are doing.

So, if you, I'm really interested to hear if you guys have a way to easily help people understand what's agentic and what is people, because that's, I think, something that's huge and something people are worried about and people are talking about. So nice teaser, Brendan. I, I'd love to tell you more on this session, but we're still about a month away, so can't really get into details.

But I can tell you, Gina, I, I'll, I promise you that if you come and listen to that, you will hear how we can give that, that traceability, that lineage, everything you wanna understand to make sure that your ag- your agents are acting safely and in a way that you can trust. So there, there will be a lot of talk about that at our, at the keynote. That's so huge, and I think for everything that gets talked about in AI, the biggest piece of a kind of the AI factory honestly is that data prep, and hardly anybody talks about it.

But, like, there's so... It's expensive. A lot of people are involved.

A lot of resources are involved. Um, and to get the data right is, is so incredibly important, and I'm glad y'all are talking about it. And that was one of the things, if I can jump in on this, that, came through last year with Qlik Connect.

We spent a lot of time talking about data quality and the impact of quality and what quality means, that it doesn't just mean, you know, that you have some data and that it's accurate data, but that you have that sort of, as you mentioned there, Brendan, lineage, that you have a sort of a quantitative assessment, a way to determine whether this data is truly high quality data. I, I'm curious if that has gone, I, I guess, out of the lab and into the, into the world, into the customer world. Is data quality a vernacular that you're hearing from your customers, and is this something that has been implemented by them?

So we're definitely starting to hear that, and I, it, it's, it's a little bit shocking to me that we're sitting here in two thousand twenty-six talking about data quality initiatives again. I figured we would've solved that twenty-five years ago. But it's coming to the forefront.

It's really interesting, and it... And I'm gonna sort of date myself and say that, you know, as, when I got into this space way back in the early two thousands, data quality was cool, and it was all the rage. You had to have the best data.

Um, two thousand twenty-six, data quality is cool again. Hashtag data quality is cool again. So what we're seeing is that customers don't wanna just make sure the data's real time anymore.

" But the added thing with agents is around how do you make sure that there's context for those agents that may make a decision off of that data? 'Cause it's more than just the veracity, the volume, and all those sort of the five V's. It really gets down to, yeah, you have your data.

It's trusted, and here's the context of what's behind it. And you know, Bryan, I think sort of you've been working with Qlik for a long time, and you know what our engine can do there. Uh, yeah, you used two words there, Brendan, I wanna pick out.

There's context, and there's trust, right? And last year, last year trust was a, a huge thing about being able to not only understand that your data has a high level of quality, but how long does it take for an organization to be able to trust their data? That's not just a data quality issue.

That is a, a human trust issue as well, and you talk about the importance of being able to interject at parts of the AI pipeline and pass it on to human intervention to be able to ensure that that trust can grow in a consistent manner. So I'm hoping we see something more in terms of those, those opportunities to be able to interject. I'm, uh...

It sounds a bit backwards, right? Because it sounds like you want to say, "I want the human back in the loop," but you still need the human in the loop at the right times, right? And that's where the key part is as far as I see it.

I can confirm you will see some of that, Bryan, on main stage and as well as in some of the breakouts. Yeah, i- isn't it funny that we're once again, once again, I, I, I feel like the, the old Bernie Sanders meme. Uh, yeah, we're, I'm once again here to, to talk about, you know, people, processes, data quality, you know, the same thing we've been talking about forever.

But it has become so much more important now because of these AI agents that people are trying to deploy because you have to figure out when people need to be in the loop. You have to figure out, the, you know, AI is both a consumer of data and a producer of data, and so you have to figure out how, what that means for your overall business processes and business intelligence and data, you know, choices. It would be very easy to, um-To lose sight of reality with AI out there confidently diving in and answering all your questions and doing your homework for you, but that's just not gonna work.

Uh, Gina, we've talked about this a lot, that, you know, people, people should be using AI as a tool, not as a surrogate. No, exactly. And you can't even use it as a tool without understanding the context of the data.

So, it's fine to wanna go fast and, and to achieve great things with automated means. Um, but especially in the case of generative AI, where it's just predicting the sequence of words, if the data is out of sequence, that is to say, not situated correctly in the, in the data set. And, like, to just to give a real world example of that, it's, it, you know, what if it's-- what if I'm trying to run a launch and the data that I'm using is from two launches ago, and I don't have all of the brand new things I wanna hype on?

Or what if my data for the market, if we're trying to figure out, well, where are we gonna, where are we gonna put these campaigns, and who should-- who's gonna really wanna buy this? Which are the first things a product marketer asks. Where is gonna-- who's gonna buy this, and where should we push on?

Like, how can you do that with data that is old? And if I go to, to do some sort of process, a-a-as I do, right, with clients, if the data I have is incomplete, which happens a lot with clients at the very beginning of things, or if it is old, then the results I give from doing something with generative AI will all be wrong, and I have to do them by hand anyway, so what's the point of using AI? Yeah.

Two things I wanna touch... No, it's great. That's, you're spot on.

There are two things I wanna touch on there. Um, just one anecdotal story. I had to write my bio for the website when I got my new job two years ago.

So I went, I did what any modern exec does, and I log into one of the generative AI tools and give my LinkedIn profile and say, write a, write a, write a job description and write a title and everything for my website. It wrote the most amazing story about me. It was amazing.

All wrong, but amazing. I mean, that's, that's the funny thing. It, it said I had my PhD in applied math and physics from Cal Berkeley, and that I work with underprivileged children on the weekends.

I wanted to post it on the website. Um, but it was... So there was that, but it brings up a more important point in today's world, right?

So it's a funny story. The challenge with some of these agentic tools and some of these generative tools is that they're very good at getting, that unstructured data in a bit of a format of a story that you can tell. But when you start adding in what has been traditionally in our space, and I'll punt this to Brian in a second here, which has been that structured data so you can make business critical decisions, the LLMs stink at math.

They're just bad at it right now. They'll get there. But this is one of the things that you'll hear about at, at Connect, where we now have the ability, and this is, this is not a pre-announce, but we already talked about this earlier this year.

We have the ability to bring that structured and unstructured data together in the same type of analysis b- And that in and of itself, that is a massive leap forward for customers who are going down this agentic path. And Brian, you've been working with us for years, so you know why we can do that. Yeah.

I mean, Brendan, I would say that, that Qlik have really been in overdrive in terms of the private preview side of things this year as well, and really getting the right access to the right customers and partners to be able to evaluate this technology as you're developing it. And I think we, we all know that the structured and unstructured data, so effectively being able to take all of the documents that you have on all of your drives across all of your network and combine it with all of that real pure data quality, build data that you have in your apps, and fusing that has been an issue for a long, long time with a lot of da- a lot of data. And I think that's what's really cool about Qlik Answers and the new version of that, is it...

And I, I, I wondered whether to think about this as a criticism or not, Brendan, but I've decided that it's not a criticism, that it actually stops and thinks, right? " And I was like, "Hold up, hold up. Do you really want to have the...

" As an, just as a, you know, a cloud-based example here. And this is the one time that you really do want to have that deep dive, and you really have to be a bit patient to wait a few more seconds to get the correct answer because, as Gina already said, if you have the wrong data coming from the right place or vice versa, and then you've got a bunch of sources, then that's when it starts to get, and I'm sure technically quite complex as well. So yeah, I think it's, I think it's a real positive that you have to wait for a little bit.

Yeah. Brian, I love the, you called it the fast button because the original name that it came up for the fast button was the Get to My, Get Me To My Incorrect Answer Really Quickly So I Can Spend Time Updating My CV, was a little bit long, right? I mean, but it's, you're, you're...

A-and Gina, you're laughing 'cause if we get this, if customers get this wrong, there are consequences. And it, it's not good consequences when, when this stuff goes bad. Yeah.

I was thinking of my own, like, so I, the tool that I'm using, I always forget to switch it when I'm doing something technical to, to the think really hard, use the newest model. And, so I did some vibe coding, and I just was trying to build a little bot to do something silly. Um, but everything on Microsoft Teams had changed.

It went to a platform since they had done the slow, the fast one. So I had to redo the whole I'm going through the steps to, to build it together, to build th-this little agent, and I'm having to read through Microsoft documentation, which is you know? So it's, it's trying to figure out.

So it, it's also just a, a human issue. So, you know, imagine doing that in a big organization and it's not, it's goingGoes how you have the easy fast button and it goes real fast and trust is totally eroded, number one, so no one wants to use it. And then secondly, number two, you're spending more time trying to get to the actions you were promised were gonna-- it was gonna do for you.

So it's, it's, it's a human issue always. It's always a human issue and, and, you know, losing that trust and making it so people don't wanna use it because it's not working correctly yet is really-- that's a cost- a costly mistake. Not to get into our technology too deep, but I think there's an-- I, I love what I'm hearing from both of you because it's-- We actually in our early access programs on our latest set of capabilities, we actually implemented a few changes to make it quicker.

" It was all wrong, but it was amazingly fast. So we actually, we, we guided Brian sort of that user experience on the grounds that people need to make decisions off of this information. More importantly, machines are gonna start making decisions off of this information, and if it takes just a, a microsecond or two seconds or three or four seconds longer, we would rather have our customers be safe and confident because that's Qlik's history.

That's what we've always done well, and we- we're going to maintain that, and you'll hear that at Connect. Big theme. So one of the things that we've been hearing about from Qlik in terms of specific technology, Brendan and Brian, and i- is this, discovery agent and, and generally in the industry.

I know Gina and I have been talking about MCP. Um, Brian, what are you looking forward to from that, and, and what are your thoughts on the way that j-- that, that Qlik is, is working in the agentic world? I, I think, I think Qlik has quite rightly eased themselves into this, right?

I think we heard a lot about the, the potential of these concepts last year. But, you know, again, with the private previews now coming through for MCP and for Discovery Agent, that word trust again is really coming to the surface, right? And it's about Qlik trusting its customers and partners to have the right level of technical understanding of what they're trying to do and to not misinterpret that.

Um, it's all great having MCP access and being able to pipe that to your favorite LLM, but if you're not educated in how to use that and you don't understand the restrictions that are there, that's, that's an important thing. But I think, I'll throw this one to Brendan as well. The Discovery Agent is one that my customers are really excited, probably most hyped about, if I'm allowed to use that word.

Um, they, they care about like all of the technical stuff, but they really want the stuff that tells them, "Oh, my, my carrots were twenty-nine percent up overnight. " Yeah, talk a little bit about Discovery Agent and why that's gonna make a difference for a lot of customers. Well, well one of the things that everybody that's listening to this should understand is that Qlik is customer zero.

Let that sink in for a second. So we are-- We use our own products, and if you think about my, my own role, Brian, as the general manager of sort of the largest division of Qlik, I want to be able to understand what's gonna happen with my business. I'll be clear, I don't want to log into my dashboard to do that.

I don't. I want it... I want an agent sitting there monitoring the health of my business, whether it's around bookings, whether it's around customer support cases, things that may even be posted to social media that can proactively tell me, "You know what?

" And that is the promise of the Discovery Agent. That's what we're really, really excited about. The second thing that we talked about, and we brought it up here, is MCP.

Uh, MCP, it's the, the latest favorite three-letter acronym that seems to be out there all the time, but it'll always be a three-letter acronym. What I'm re-- what we're really excited and what our customers are telling u-us they're excited about is that MCP is going to allow our customers to bring the power of our trusted data products plus the engine to where users work. If you work within Claude, and that's the way you run your day.

Gina, you talked a little bit about the product marketing world. Probably spending your time in Perplexity or Claude to try to drive some things of what's happening there. Whether you use Perplexity, Claude, Gemini, I mean, we can continue the list, it goes on forever and ever.

MCP is going to allow all of our customers to bring that power into that user experience, and that is gonna fundamentally change the way people engage with data. And technically, why that's so exciting is the engine, our analytics engine, has the ability to ensure that the data within those questions you ask through MCP, through Claude or anything else, that they're accurate and they don't hallucinate. That is what is extremely exciting because I challenge anybody who's listening right now to dump a spreadsheet into any one of the LLMs.

Actually do, do something this way. Dump five or six spreadsheets in and then start asking questions and have it build you a story. It'll be really pretty, but it'll be really wrong.

That's what's so exciting about MCP. We can address that problem, and the promise is there. We've delivered.

We won't just do a nice pretty video on stage. I happen to know we'll be able to do something live and in person with, with real use cases as early as next month. Really exciting times.

Yeah, look, I think just, just to sort of tail off on that for y- for you, Brendan, I think that use case part is super important, right? Um, I think g-getting past the hype phrases of Discovery Agent, MCP, Claude, I think a lot of people in the market who even have some like touchpoint there know about those things and that they exist. But I think the use cases are what, what's really gonna make this for the customers right now.

That's what they want to see. How can I practically use this in my business? I think you used that same phrase right at the start of the, the session, Brendan, so Sometimes technologists get enamored with technology for technology's sake, and it-the- it's cool, right?

There's nothing wrong with that. " They have to be able to go with real use cases where they can apply it, not nine months from now, not nine weeks from now, heck, nine days from now, they need to be able to do that. And that's the promise that we want look to deliver against here or within over the coming months 'cause there's more than just the discovery agent.

So stay tuned to Qlik Connect. There's more coming. That's what we're really excited about.

You know, one of the things I like to Qlik Connect too, is your users force you to talk about what you announced last year . And I, I remember last year the, the different user, presentations that were there were really good because they came up with, "Yeah, this was announced. Here's how we used it.

" So, I'm hoping to see some of that again this year too. Well, that's actually a great question. Um, you know, let's, let's kinda transition here into what we're looking forward to with Qlik Connect.

I mean, you know, for me, it is what Brendan was just talking about, these kind of real world use cases. I mean, that's, that's where the gold is, when you get to sit down a-across the table from somebody who is implementing this stuff and hear how they're doing it, what they really think of it, you know, unfiltered. Uh, you know, this is the real state of affairs.

But of course, there's also a lot of oth- other things that happen. You know, we've got a couple of, keynotes. Um, I, I'm hearing rumors that, one of them might be especially amusing, the second keynote there.

Uh, the first keynote's gonna be all, technical, and we're gonna be live streaming, our, our feedback from that on the, on the Tech Field Day site. We're gonna be, doing a live blog, I should say, on that. Um, Gina, you were at Qlik Connect, with us, previously.

What, what are you looking forward to? What do you li-like to see again? Uh, everything is good.

Um, what I like-- One of the things I did like, honestly, is the, the, the customers showing off what they have done and how they've been using it. So you hear the customer use cases. And, and all the community events.

Their community is super activated and not shy in any stretch of the imagination. So they're very happy to tell you what's going on and what they like and what they don't like. Um, I think that's pretty much what I'm looking forward to.

Yeah, yeah. For me, it's, it's the energy on the floor. Um, I think you can go to many tech conferences, and you just don't have that.

Um, and that's, that's twofold, really. That's the, the customer energy that's on the floor, the, the buzz that's there. But I think, I think it's also fair to say that the, the Qlik exec team do a really excellent job of being around, and I mean around in the truest sense, right?

Where they're there to mingle where they can, possible, and it's, it's really nice to have that. It's a different kind of vibe from going into, a gray square and walking around a bunch of booths. Qlik Connect isn't really like that, or never has been in my experience.

So I'm really looking forward to seeing some of-- 'cause we always get some surprises, so I'm looking forward to what's, what's around the corner this time. And Brendan, you get a different perspective on the show, but, what's your perspective? Yeah.

I'm looking at all the insider trading notes I ha- I have on here about what we're gonna announce. Um, I'd love to share it on this. I can't.

Um, my perspective is, so I was unable to atten-attend Qlik Connect last year 'cause I had a, a bit of a health crisis, so I'm actually just looking forward to going back, and seeing some of the customers, that I know and, and really respect. I think, Gina, you said it well. Some of the customer stories, they're surprising at how much of an impact that we really have as an organization, but it's where we've already always tried to be successful and help our customers.

And Brian, just you've been in the ecosystem for years. I'll just say I love watching some of the partners show up dressed in full-on green. Um, you know, whether it's the green shoes, the green so-socks, or, Dalton, if you're listening to this, your green tie, we love that.

Um, I, I can promise you, I will have my, my green Sambas this year, so we'll, we'll really get into it. Um, but ultimately, it gets back down to talking to customers and the partners. That's, that's what really, really blows my hair back on this.

So I'm looking forward to it, and then the surprises that we're about to announce in a couple weeks here. And, and, you know, and one final thing too that I really enjoy about, Qlik Connect is the sort of the vibe on the expo floor. Um, sometimes, as Gina and I have experienced, for many years, and Brian, I know, you know, you and I experience this as well, at some conferences, it's a little lame.

It's a little boring on the expo floor. Uh- Yeah ... that is not how the Qlik Connect expo floor is.

Uh, between the, the silent disco headphones, there's constant presentations. There's, sure, there's partner booths and Qlik booths and so on, but it's very demo-heavy. It's very inviting.

And also, there's always some other stuff on there, you know, whether it is bicycles or motorcycles or race cars or whatever. You know, there's always something fun and, and different and interesting to see, and so I'm really looking forward to seeing that as well. And if you are going to Qlik Connect, please do stop by.

We will be-- The Tech Field Day organization is gonna be doing a lot of stuff in the, the media area at the center of the expo floor. We would love to talk to you. A-also, if you're, if you're listening to this and you're saying, "Well, this kinda sounds kinda cool," you can still probably get there.

Um, I bet if you go to the Qlik Connect website, you could probably still, get yourself to Orlando and, and join us. Uh, there's still time, so please doCheck that out. Um, before we go, I just wanna give each of you a chance to tell us where can we continue the conversation, where can we connect with you.

Uh, Brendan? com. That is our main website, but I'd also encourage everybody that's listening in to connect with me on LinkedIn.

Love hearing new perspectives. Yeah, same for me. I'd, I'd love to meet as many of you as, as possible at Qlik Connect this year, but you can, you can mostly find me on LinkedIn.

That's where I live, so... Yeah, you can, find me on LinkedIn as well, but I definitely wanna hear the tea. That's what our podcast is all about, so, check in to Tech Aunties podcast, and if you wanna talk to me live while we're there, I would love to have that opportunity.

Yep, and you'll find me, sometimes here on the Tech Field Day podcast. Also, every Monday live with Techstrong Gang, which is a daily news program, as well as, Wednesday, on the Utilizing AI podcast, which I host, which is very much in line with our conversation here in that it's all about practical uses of AI rather than the hype that we're constantly living in. So thank you both, all of you for, delegates, Brian and Gina, you know, Qlik, Qlik, Brendan, thank you for joining us, and thank you everybody else for listening to this episode of the Tech Field Day podcast.

If you enjoyed this, please do subscribe. You can find this on YouTube, with video if you'd like to see our lovely faces, or of course, you can also subscribe in your favorite podcast application. Uh, this episode is brought to you by Qlik and Tech Field Day, and also, of course, the Futurum Group.

com/podcast, or you can view us and learn more at Techstrong TV in your favorite streaming device. Thanks for listening, and we'll catch you next week.