Pragmatic Works is a Two Word Sentence, w/ Brian Knight

Rob Collie

Founder and CEO Connect with Rob on LinkedIn

Justin Mannhardt

Chief Customer Officer Connect with Justin on LinkedIn

Pragmatic Works is a Two Word Sentence, w/ Brian Knight

In this episode, Brian Knight, founder of Pragmatic Works, describes his accidental entrance into the field of technology. From a chance job offer to being “that guy” who hit the server rebuild button, you won’t want to miss this data origin story.  Only someone like Brian could get such a great start by nuking a database and turning it into a DBA career where he would become both a prolific community member and content creator.

While reminiscing with Tom about SQL and the glory days of being a DBA, Brian admits that his focus has recently shifted to Power Apps. When asked why he shares that the lure of building in a couple of days what in the past would have taken several years was just too strong.

It isn’t just about the past, though, as the crew gets into a lively discussion on AI and its future impact on the arts. No spoilers here but it may just be that AI can be a beneficial part of creative society, not by replacing people as artists but by augmenting the existing skillset of current and future artists in a similar way the Power Platform enhanced skills of people in tech to create the Citizen Developers we all know and love.

If you enjoyed this episode, be sure to leave us a review on your favorite podcast platform to help others find us.

Also on this episode:

SQL Server Central

Tom’s blog: Operations Manager and SQL Server

Septic Companions and Breaded Capacitors, w/ Chris Rae

GitHub Copilot

Open AI

Two QB Rob Blog

Cat in the Hat: Calculatus Eliminatus

Episode Transcript

Rob Collie (00:00:00): Hello, friends, and welcome to 2023. We took a little bit of a break over the holidays, but we are now back. And to kick off the year, we welcome Brian Knight to the show. Now, Brian is someone I've wanted to have on the show for a long time, but there's a little bit of an awkwardness about inviting the founder and CEO of a competitive firm, one of our competitors onto the podcast, right? What do you talk about if you invite a competitor? Is it going to be impolite? Is it going to be awkward? Well, Brian wanted to be on the show so bad that he went and sold his company. Nah, I'm kidding. Anyway, so Pragmatic Works. You almost certainly know, dear listener, you almost certainly know the name Pragmatic Works. They're at the forefront of training on the Power platform. And so yeah, they are still a competitor of ours.

(00:00:45): Training is a part of our business at P3. It's a fraction of our business, but it's enough of a fraction that I think Brian and Pragmatic Works are now below the awkward podcast guest threshold. And yes, I have wanted to talk to Brian on this show for a very, very, very long time. He's always been just so energetic, so sincere, so dedicated and passionate about what he does, and also incredibly warm, friendly, and generous. You'll absolutely hear that during our conversation. You'll also get a sense of just how humble he is. And that, dear friends, is something very, very difficult to do in life. Is to be essentially completely on top of the world, on top of your game, and at the same time remaining humble. We recorded this before the holidays, so it's been long enough now that even I am looking forward to hearing what we talked about. Are you excited? I'm excited, so let's get into it.

Announcer (00:01:41): Ladies and gentlemen, may I have your attention, please?

Announcer (00:01:45): This is The Raw Data by P3 Adaptive Podcast with your host, Rob Collie and your co-host Thomas La Rock. Find out what the experts at P3 Adaptive can do for your business. Just go to P3Adaptive.com. Raw data by P3 Adaptive is Data with the Human element.

Rob Collie (00:02:11): Welcome to the show, Brian Knight. Long time coming. Really, really happy to have you here.

Brian Knight (00:02:16): Thanks for inviting me.

Rob Collie (00:02:17): Are you on a treadmill desk?

Brian Knight (00:02:19): I am. It's my little squirrely thing that I do now. I've got wicked ADHD, so it kind of helps hone my ADHD.

Rob Collie (00:02:25): This further confirms my instinct from a distance that you and I really should have hung out much more than this. Because believe it or not, the previously released episode, our guest was also on a treadmill desk.

Brian Knight (00:02:38): Oh, really?

Rob Collie (00:02:39): And we talked about this and my observations from when I had a treadmill desk that it did nothing for my physical health. However, it calmed my ADD.

Brian Knight (00:02:47): It does.

Rob Collie (00:02:48): I was so much more productive.

Brian Knight (00:02:50): I can actually stand to write a paragraph of text without opening up YouTube or something. And it does help a lot. From my case, I think it's just our company walking during the day and we have a neighborhood right near us, we'll kind of walk around the neighborhood. I'm sure the people in the neighborhood think we're complete psychos, but we always bring candy with us to give to the kids as well.

Rob Collie (00:03:09): Oh yeah.

Brian Knight (00:03:09): Just in case.

Rob Collie (00:03:10): The white panel van.

Brian Knight (00:03:11): Exactly.

Rob Collie (00:03:12): Right, yeah. But you have the panel van riding along, kind of cruising, idling along next to you as you walk, tight?

Brian Knight (00:03:19): Exactly.

Thomas LaRock (00:03:19): Other than Halloween, who goes around the neighborhood with candy?

Rob Collie (00:03:24): He's kidding, Tom.

Brian Knight (00:03:25): What kind of greedy jerk are you, Tom? Come on.

Rob Collie (00:03:27): Yeah, Tom just goes around the neighborhood handing out money. It's like, "Here, peasants, takes some."

Brian Knight (00:03:34): But I have noticed my coding gets better. It's so much for our time is spent just sitting down. It definitely breaks the day up a lot more also. Although my feet really, really hurt the first few weeks since you're doing it.

Rob Collie (00:03:45): Yeah, this is terrible. I might need to go get another one. I might need to get a treadmill desk again. This desk that I'm sitting at, it elevates. So, I already have half of the story, the standing desk. Have you burned any treadmill desks out?

Brian Knight (00:03:59): Not yet. This is only about six months old. For me personally, I lost about 10 pounds on it in the first eight weeks or so. So for me it was a big deal. I think I was going the wrong direction. Being a tech professional, very easy to start to put on some pounds and see your baseline start to bump up a little more each time.

Rob Collie (00:04:17): This morning, I commemorated a crucial just absolutely ground shattering milestone of my 500th Orangetheory fitness class. And because it's been tracking my heart rate the whole time and knows how heavy I am, it tells me that I've lifetime at Orangetheory Fitness, I've burned like 153,000 calories.

Brian Knight (00:04:37): Holy cow.

Rob Collie (00:04:40): And so this equates to what, more than 40 pounds? No, not more than 40 pounds. Math is hard, but it's definitely north of 30 pounds. And I actually think I eat less when I'm exercising there. So 40 pounds of fat is the floor for how much I have not put on.

Brian Knight (00:04:56): That's awesome. I did Orangetheory for about six months also. It's a great program. Until you go get sick or you go on vacation, then you kind of get the habit of doing it, then it's so hard to get back in the habit, so.

Rob Collie (00:05:06): Yeah, I try to go in the morning before my brain really realizes what's coming for it. I'm walking in the door going, "Oh."

Brian Knight (00:05:13): But did you puke the first time?

Rob Collie (00:05:15): No, no. I have had to leave the class a few times as I started to experience some feeling of nausea and I introduced someone to it who then threw up after the class one time. I'm like, puke adjacent with the Orangetheory Fitness. But yeah, treadmill desk. The ADD. ESPN is calling. YouTube is calling. Twitter is calling. There's so many things, they're just begging for your attention. Bet there's something good on Reddit.

Brian Knight (00:05:42): You got to do 20, right? It's just, it's given.

Rob Collie (00:05:46): Yeah, this is serendipitous. I think the universe is trying to tell me I need another treadmill desk.

Brian Knight (00:05:50): Absolutely is.

Rob Collie (00:05:52): Our last guest just kept walking out the whole time we recorded.

Brian Knight (00:05:54): I had somebody before who sent me a text message on my meeting. He's like, "Hey, could you go ahead and stop your camera? You're making me sick by the whole bobbing motion." So, I keep it pretty low when I'm on camera so nobody really notices, usually.

Rob Collie (00:06:06): Knowing what I know now, I would take you walking on the treadmill desk as a sign of respect.

Brian Knight (00:06:11): There we go.

Rob Collie (00:06:12): You want to provide us with your undivided attention. Now that you're standing still, we're going to be wondering, "Is he on YouTube? Is he knuckling down and giving it-

Brian Knight (00:06:20): There you go.

Rob Collie (00:06:21): We want to get the whole story. We want to know Brian by the time we're done. Where do you want to start? I'll tell you what, I'll give you, choose your adventure.

Brian Knight (00:06:28): Oh, okay.

Rob Collie (00:06:29): Okay. Start with today and work backwards. Or we could start with in the yesterday year, in the way back machine and work our way forward. And either way we're going to jump around a lot. Where would you prefer to start?

Brian Knight (00:06:39): Squirrel. Oh-

Rob Collie (00:06:40): Of course. Yeah.

Brian Knight (00:06:42): Yeah. I owned a company called Pragmatic Works, and we used to be three companies, training, consulting, and software all around data and analytics and Azure, as well as now we're into the app building with Power Apps. But when I first started, I started in the .com era where anybody that said the word computer could get a job. And I was working as the translator at AT&T at the time, and I walked in to get my first internet CD. Those AOL CDs that we used to have, well, the local providers had those also. So gosh, we're all that same age, aren't we, this sucks. So, walked into the local internet provider and the guy who owned the company had all these wires hanging down. He was all stressed out. His tech support person just quit, probably fortuitous, probably should not have done that. But I went ahead and he offered me a job right there, found out I was working at AT&T.

(00:07:30): So that was my first job in the tech field back, I was probably 19 or 18 at time. That turned into doing building websites, which eventually after they got acquired and they got acquired, we became a pretty good size company providing internet service around the community. Back when anybody can start up an ISP.

Rob Collie (00:07:48): Yeah.

Brian Knight (00:07:48): I walked in the door one morning and the database was down. I kind of knew what databases did back then, but not really. And this is the .com era. I had 22 people working for me, and a lot of them were on the younger side and the less professional side. And so my DBA was constantly coming to work with a little red eyes, you understand what happened there. He had some red eyes and couldn't get a hold of him. I called him, called him, called him, there were hundreds of customers down. I sent somebody over to his house and knocked on his door, and he kind of came the door all staggering and all right, this isn't going to happen. So, I pull up SQL Server and it said somewhere, "Repair the database." That sounds like a good idea, that's exactly what I want to do. I want to repair this database. So, I hit that and it said, "Rebuild master database." Yeah, let's do that also.

Rob Collie (00:08:36): Yeah, this button says, "Fix it." Yeah, I'm going to press that button hard.

Brian Knight (00:08:39): And of course, right when I did that, the DBA walked in and his red eyes immediately went away and he goes, "What did you do?" And it came up, SQL Servers up, see, it's working now. I can actually open up this little tool now. And he goes, "Yeah, but all my databases were gone." This is for, we had NFL teams right around the server. It was the late '90s, so it was much more decentralized. It wasn't a Rackspace, it wasn't an Azure. It was all decentralized at that point. So, we spent until about two or three in the morning kind of rebuilding that database. And I'm like, "I want to learn some stuff about this." When I made that next job hop, I'm going to learn everything I can about this and become a DBA.

(00:09:19): I did two interviews that day. One went remarkably well because they not asked me a single technical question the entire time. This is more of a personality interview. Got the job on the call on the way home saying, "Hey, you got the job." "Great, I had one more interview to do and I'll call you tomorrow." I did that interview and they started to ask me technical questions and I was, there's some pretty humbling experiences we all had in our life, but that was a crash and burn hardcore. They're like, "I like you, Brian, but you suck." So, it was one of those things that I was like, "Maybe I'm not meant for this." But luckily that first job, that first database job was so slow as far as what they had to do, I had a great opportunity to learn. I learned best by teaching and kind of writing and all that. There's a consultant that was sitting next to me and he's like, "Hey, I'm writing his book. Do you want to join me?" And I'm like, "Okay."

(00:10:08): I was young and stupid. And so I started writing with him and I realized how much I love that. And I started working on a website to promote the book called SQL Service Central Now. So we built that website at the time to promote the book, and then later it did a whole bunch of other good stuff also. That's kind of how I started in the data industry.

Rob Collie (00:10:26): So, the on-ramp through the ISP world, that's a new one, really every one's on ramp is different. I love that, it was sort of like you were already in tech and then you made this hop from tech to tech. I'm also aware of that Wild West early era of ISPs. Because there was one in Nashville called Telelink that had been founded by, I think a couple of Vanderbilt grads. The servers were in their office.

Brian Knight (00:10:52): Oh yeah.

Rob Collie (00:10:53): It was pretty loose, but it was the best internet available in Nashville at the time.

Brian Knight (00:10:58): This is one of the only places I've ever worked where I had to let somebody go for installing Doom on a server, on a production server.

Rob Collie (00:11:04): Yeah.

Brian Knight (00:11:05): Well, it had the best internet. Yeah, but-

Rob Collie (00:11:08): Yes, it does. It absolutely does.

Brian Knight (00:11:12): For that I started down in the BBS land, I'm not sure you remember the old BBS term back then, Bolton Board Services, those were the precursor to the internet. And I was 16 and had 12 phone lines coming to my house. My parents loved it. They loved it. Because I was up until two in the morning all trying to rewire stuff and all that. I'm always struggling with the curiosity aspect. When you want to hire somebody, how do you hire for that kind of curiosity?

Rob Collie (00:11:35): It's hard to recognize, but it's also crucial. A big chunk of what I call the data gene is curiosity. So, SQL Server Central, I am familiar with this domain name I'm familiar with this website, had no idea this had anything to do with you. What did that turn into over time? And Tom, do you have any fond memories of SQL Server Central, which by the way, major points for the alliteration in that name is SQL Server Central still going?

Brian Knight (00:12:01): It is, yeah. We ran it for, rubbing on the side. I would work all day and then generally do about four to six hours at night, sending out a newsletter every single day, writing an article every day or whatnot to keep people interested in this website. When we first started it, there was only a few players out there at the time for community websites around SQL Service before everybody had their own blog and all that. We eventually sold it six, seven years after we started it to Redgate. And at that point they grew it even farther. We grew it to about 800,000 people that were registered and then they've grown it to somewhere in the millions at this point.

Rob Collie (00:12:37): Now, Redgate, is this a competitor to you, Tom? I forget.

Thomas LaRock (00:12:41): My standard answer is simply no, because my company really has no competition when it comes to what we do.

Rob Collie (00:12:48): Sure, sure. I mean, they don't have a Tom, I mean-

Thomas LaRock (00:12:50): Well, no they don't. But in Redgate's case, they actually do have a Tom. Redgate is very much, they've always been more on the developer side of things, that side of the house. And they dabble, say in database monitoring and performance and all that. But even today Redgate would tell you that they are a Dev Ops focused company, that is really their mission. And they aren't doing enterprise infrastructure monitoring, which is what we do.

Brian Knight (00:13:15): As I mentioned before, SQL Server Central really started as basically a personal blog at the time. And there was a bunch of the authors that were writing for this website called [inaudible 00:13:25]. And so we decided to band together and then build this website. We would candlelight, work on it after hours and eventually we got it up. It was pretty horrendous looking, but then we eventually iterated into a pretty awesome website I thought. And then Redgate bought it about six or seven years after we started it and really advanced it from there.

Rob Collie (00:13:45): Did you say tried to write one newsletter a day?

Thomas LaRock (00:13:48): He did, yeah.

Brian Knight (00:13:49): It was, and it had to have some unique content, some kind of editorial kind of content in there also. So, my wife was not very happy for a while. And then eventually Steve Jones, you likely know also, he started running it full-time for us versus kind of moonlighting what we were doing. And Steve Jones is already a partner, so he just rolled him into running it full-time for us and he did an amazing job.

Rob Collie (00:14:13): As someone who used to write two blog posts a week and thought that was killing it in terms of volume-

Brian Knight (00:14:20): It is, yeah.

Rob Collie (00:14:22): One a day, are we're talking seven days a week or-

Brian Knight (00:14:25): Five days.

Rob Collie (00:14:26): Five days, but still crazy. That's a lot of output. How long did you sustain roughly that pace?

Brian Knight (00:14:32): The moonlighting for several years. And then Steve did it, but he did it full-time and he was able to keep it, escalate even better. So he made the good, the great at that point.

Rob Collie (00:14:40): Tom, are you thinking of something there?

Thomas LaRock (00:14:42): Yeah, I was just going to say, around the time that Brian, Steve and Andy were doing their stuff at SQL Server Central, those are my formative years as a production DBA. SQL Server Central is where I had my first technical blog post published.

Brian Knight (00:15:01): It's still up.

Thomas LaRock (00:15:02): I had met a handful of people through past summit and one of them was this woman named Kathy Kellenberger, and she agreed to do a tech edit of the post I wanted to submit. And we sent it off to Steve and he was more than happy to accept it. At the time I thought it was a big deal, like, "Hey, I got something published, this is great." Now I realize, "Well shit, they'll take content from anybody." So they're starving for content, they'll take whatever, it doesn't matter. But anyway, it was a really good.

Rob Collie (00:15:32): Now.

Thomas LaRock (00:15:33): No, it's true. It's a really good introduction for me in an aspect of technical writing. Yeah, my first article, it's still up there and it was a painful moment for me. And it was a little learning moment for me that I turned into that post, which was, I think, are your backups online and available? Because the way we were storing our backups, I could actually go and check and say, "Okay, the file is still on disk and can I recover it or is it readable?" And it was just a process I put together out of necessity. Because we would have backups suddenly disappear because the server team just figured, "Oh, I don't need those four backups, they can all just go away." Without understanding the chain of a full of differential and transaction logs. They were like, "That file's large and it's three days old, I'll just delete that." And I'm like, "That's a problem. I won't be able to recover from that." I had to write it and that became my first post.

Brian Knight (00:16:31): That's awesome, Tom.

Rob Collie (00:16:32): So the post wasn't painful?

Thomas LaRock (00:16:34): It was me not being able to recover a database. I have one job, two jobs. I need to reset a password and I need to restore a database. Congratulations, you're a DBA. And if you can't recover, you can't keep your job. If somebody comes to me and they say, "Oh, I need you to put us back to four days ago, we have to research an issue." And I'm like, "Yeah, not a problem." I go to do it and the files aren't there. Now all of a sudden I'm in trouble and I didn't do shit. I did everything I was supposed to do, but somebody in the server team decided to remove a file and they're like, "No big deal. We'll just go recover from tape." "Okay, great. How long's that going to take?" "Well, you'll have it here in a few days." "Can we do it sooner?" "Well, if we pay for the emergency, they can get it here in a few hours."

(00:17:16): So what do we want to do now? At the best, what should take five minutes I now have to tell the business is going to take a few hours if they're willing to pay, otherwise it's going to be a few days. And again, now I'm the focal point of their anger and I didn't do shit. But I can mitigate my risk for next time by checking to make sure that my backups are online and available.

Rob Collie (00:17:40): By the way, anyone out there listening who wants to learn mitigation strategies from when they're being blamed for a failure, "I didn't do shit," is not a good thing to say. It's just begging for the retort, which is, "Yes, exactly. We needed you to do some shit."

Thomas LaRock (00:17:58): But it wasn't my mistake. All my stuff's working. Somebody else decides, and it's, yeah, it wasn't my fault, but this is way I learned actually. I had a great manager who basically said at the time, he goes, "Oh, it wasn't your fault, but it is your responsibility." And I'm like, "Ugh, yes."

Rob Collie (00:18:16): Yeah, see that's good. That's good. That's good. That's what we like to hear. I was going to say, sitting next to the engineering professor on an airplane flight one time. Who was telling me that there was no engineering failures at Fukushima, because it had been designed to survive a 100-year tsunami. What had happened was a 200-year tsunami. He was technically correct what he was saying, but the lack of any additional context, that was the end of the sentence, the period, see, no problem. Right. I'm like-

Thomas LaRock (00:18:49): Meanwhile I'm picturing that fish in the Simpsons with the three eyes just.

Rob Collie (00:18:53): Yeah.

Thomas LaRock (00:18:53): Yeah, no problem. He's on the side of the river like this. No, no. We did everything we could.

Rob Collie (00:19:00): So there we go. Bringing this back to Brian. Brian, you are responsible for the phenomena, the internet phenomena known as SQL Rockstar, because your organization was the first one to put a microphone in his hand.

Brian Knight (00:19:14): It's true.

Rob Collie (00:19:15): He's like, "Oh my God, I love this. I'm not giving it back. Piss off, this is my microphone."

Brian Knight (00:19:20): Well, and Tom will attest to this, right? Back then there wasn't a lot of individual blogs. There was Penton with SQL Server Magazine, there was a few other sites and that was pretty much it. So, the credibility that you got for having something published is more credible than some dude walking in with a five year or six year degree saying, "Hey, I've got this paper right here, I'm ready to go boss." Back then, I think personally that that experience outweighs even things like college back then. From a college dropout, let me rephrase that.

Rob Collie (00:19:49): Well, America has been built by college dropout. There's Bill Gates, there's, did Jobs ever even go to college?

Thomas LaRock (00:19:56): Technically he went, he didn't finish.

Rob Collie (00:19:58): Okay, okay. And then there's this other company I've been listening to all these podcasts about lately, this Theranos company, impactful corporation founded by a Stanford dropout. My wife and I are just eating this up. We're consuming it on 18 different channels.

Brian Knight (00:20:14): [inaudible 00:20:14] Series now too.

Rob Collie (00:20:15): So, what year was this first blog post on SQL Server Central?

Brian Knight (00:20:19): Oh gosh, I'm going to say 2005 or '06. I can look it up.

Rob Collie (00:20:24): So yes, I can attest, the blogosphere was sparse. And Brian, I liked what you were saying about credentials and credibility. The magnitude that was brought by saying, "Hey, here's my published post." It's the equivalent of being published in Nature Magazine or the Nature Journal. There weren't any peer-reviewed journals. But this essentially was peer reviewed and Tom said like, "Oh, it didn't mean anything. They'll take anything." That's not true. This is how imposter syndrome works, is you just constantly reset your zero. And this is good for progress. It sort of perpetually drives you forward. Probably not so good for the psyche.

Brian Knight (00:21:02): And it's tougher nowadays, I think, because it is so decentralized now and I've written in the past 18 books. Doing it now, I'm not sure I would do it again because the technology changes so much and people don't read books like they used to. So, it's tough the next generation to kind of get noticed now I imagine from that credibility side. It's going to take so much more work than what I had to do in the past.

Rob Collie (00:21:25): Completely, even relative to me starting my blog in 2010. I had the benefit of jumping into a brand new technology at the time. That also had some unwarranted and unjustified credibility issues with the established guard. So, I got to run unopposed in my DAX blogging.

Brian Knight (00:21:42): Was Tom pulling you again?

Rob Collie (00:21:44): No, no. Tom met me the first day and said, "Okay, I'm taller than Rob, but he'd probably be a tough out, so I'm going to go find someone smaller and weaker."

Thomas LaRock (00:21:53): It's not usually how I operate. I found the article, by the way, it's from March of 2006, so.

Brian Knight (00:22:01): Sweet.

Thomas LaRock (00:22:01): Over 16-

Brian Knight (00:22:02): That's awesome.

Thomas LaRock (00:22:03): 16 years ago.

Brian Knight (00:22:04): It's back when I had air. Yeah.

Rob Collie (00:22:05): All right. So how does Pragmatic Works come to be?

Brian Knight (00:22:09): Yeah, so fast forward many, many years, and I was a consultant at the time. And one of the projects I was on was replacement of, it's old things, you may not know the term, but it was called DTS, Data Transformation Services that was built in this SQL server that was for loading data warehouses and loading reporting systems usually. I was on a project where we were converting of that to SSIS, the new tool. And I'm like, there's got to be a better way of doing this. I started a company using my SQL Server Central sale to basically build that tool to convert DTS to SSIS. Microsoft had a path that was really lackluster, so we through several reboots and attempts and tries we eventually built the application.

(00:22:50): Took a long time, a lot longer than I thought it would take to build it. We release it and nobody downloaded it. Much less nobody bought it, of course. And we were using consulting to help fund our development efforts and I would consult for 80% of the time and do this 20% of the time. All the money went to pay a development cost of this application. The same journey you went through, I imagine. Eventually we hired my first partner, salesperson, and he was able to, we did a lot of webinars and all that and all of a sudden we went from that first week when he was on with this webinar series. We actually broke good webinar at time. We took him down all week, they kind of calling us, "What are you guys doing?" Eventually that increased our downloads and we were able to start to enter our empathy and get forward. But we were on that, "Do we keep doing this? Do we not keep doing this," early on for quite some time until we built that Salesforce team.

Rob Collie (00:23:40): So, is that when the name Pragmatic Works was chosen.

Brian Knight (00:23:44): We mentioned before how SQL Server Central, how you can actually get a decent domain name back then. When I started Pragmatic Works, I just had a bottle of wine and then GoDaddy. I know I want to have works in it and I'll just keep on throwing a whole bunch of adjectives and nouns until I finally hit something. And then with that and a $49 logo from a logo maker company, we are ready to rock and roll.

Rob Collie (00:24:04): There you go. That's all you need for a company.

Brian Knight (00:24:06): Exactly.

Rob Collie (00:24:06): You got the domain name, you got a vague idea and a $50 logo, which probably cost you closer to a 100 today, but that's it. That's a business plan.

Brian Knight (00:24:15): Exactly.

Rob Collie (00:24:16): Had to have works in the title, the word works in the sense of skunk works, like a factory, that kind of works, or this shit works. Of course you like both. You like both.

Brian Knight (00:24:28): It doesn't matter I think in my case, but yes, I like both. But the latter was really the goal. When we first started consulting, one of my partners at the time was like, "Oh, they want to build a warehouse. It's going to take us years to get that done. Well, what if we can do it this fast? How can we do that?" We've wanted to find a ways of breaking the mold of the big four consulting companies that are out there and do things a little more efficiently. They may not be as polished, but we're done in one quarter of a time potentially, or one 10th a time. That's why I wanted to have that works in there is being more pragmatic about getting something to production.

Rob Collie (00:24:59): All right, so is it like a two word sentence?

Brian Knight (00:25:03): We'll go with that.

Rob Collie (00:25:03): Pragmatic Works, is works a verb? I've always liked the name, but you can do so many different things with it in your own head. Is that how you thought of it? The pragmatic approach works?

Brian Knight (00:25:16): That was a goal. Whether it resonates, how to spell it for most people when I'm telling them the company name and whether it works or not, we have to determine. But so far so good.

Rob Collie (00:25:23): All right, so there you go. There's the origin story for the name Pragmatic Works. I've always thought of it as works as a noun, as the factory or the foundry, but it's a pragmatic flavor of that. It was still very compelling to me, even though I had it wrong, Brian. I had it wrong.

Brian Knight (00:25:42): Yeah, no, it's all good. And it's part of that, we all had our journey to get into technology in general. That aha moment, you mentioned the data gene, right? I felt frankly you mentioned imposter syndrome also. I felt very much that way early on and very blessed. I use that verb in my case, very blessed that I got here. And back when we first started the company in 2008, there was about the worst possible time to start a company. The economy had hit the gutter. But then the unemployment rate in technology was like 3%, but nationwide was like eight or 9% unemployment rate. And then we had these veterans that were coming back from Iraq and Afghanistan who could not transition well into civilian world. So, we built a foundation at the time to get those that were veterans, unemployed people and underemployed people, people making less than 45 at the time, now it's closer to 50, into IT.

(00:26:39): And worked out really cool because it was finding these people that were super, super passionate that we're willing to learn and grow and they're willing to put the hours in, because they were veterans, and getting them into IT. And we had some good community partners around town that would hire these folks. So we put them through a week long bootcamp, put into an apprenticeship and then usually that we find some skill in the SQL server world that was a good entry level skill, and that was reporting service at that time. Building reports for communities and companies was a great place, great starting point. And then from there you could go into data warehousing and wherever else.

Rob Collie (00:27:12): I love the metaphor here, which is that returning from war, struggling to readapt to civilian life, so we introduced them to SQL server reporting services instead. As an alternative to civilian life.

Brian Knight (00:27:29): We had folks that were lifting boxes at UPS, or lifting guns in Iraq and now about 500 of them and those first five years have found employment and there's a lot of these folks now are working at Microsoft, working at these big enterprises now making six figure where they were making the 20s and 30s before, or unemployed before. Personally I felt that way because I was so fortunate to walk in that ISP that day and saw that guy with his wires and he just happened to offer me a job. I have a big debt to pay because I would not be here with those four or five forks that we had in the road.

Rob Collie (00:28:01): Yeah, the fortunate bounces.

Brian Knight (00:28:03): Yeah.

Rob Collie (00:28:03): I mean, you look back though, there's also a whole bunch of unfortunate bounces as well. You know, you've got to sort of steer the bounce just a little bit.

Brian Knight (00:28:13): I mean, I think in our lives you get four or five really, really big forks in the road. And if you make the right decision or just happen to fall in the right decision on those forks, then your life can go a whole different direction.

Rob Collie (00:28:25): Well, this is something I've always appreciated about you, mostly from afar, but it's amazing how clear an impression you can get sometimes on a very small amount of interaction. From a distance I've always thought of you as definitely a professional and a technologist, but very, very human. That stands in pretty sharp contrast to most firms in this space. They almost strive to be faceless in a way. The big four, almost criminal.

Brian Knight (00:28:56): Suffocating for its employees.

Rob Collie (00:28:57): Suffocating, yeah. We're making enemies left and right here. Let's focus, sit on the positive, not how inhuman the other people are. Let's focus on how human Brian has always been. You were doing the ADD thing of three different business models, somewhat related for sure. You mentioned them, right? There was consulting, software, what was the third one?

Brian Knight (00:29:21): And training. Yeah.

Rob Collie (00:29:21): And training. Yep. All right. And I always thought of you over the years as primarily you were the one who was into the software part. Was I right about that?

Brian Knight (00:29:30): It was. So, every time we sold a company, we sold software about six years ago and it actually, it's under Tom's belt now. I think it's a SolarWinds company now at this point. It went from there to CenturyOne and to SolarWinds, I had to kind of relearn who I wanted to be in life. You go this period of depression almost as you sell a company where you're like, "What do I do now? What am I going to do with my life? I have this big non-compete that says I can't do X, Y and Z, what I've loved for the past 20 years." And you have to reinvent yourself every time. When we sold Consulting about three years ago, my reinvention was to go more into the Power app side and do more of that side of the house of the Power platform.

Rob Collie (00:30:10): Training primarily, or implementation? What's the mix there?

Brian Knight (00:30:14): For what now? It's all training a 100%. It's so much more fun to enable others than to have to get that call on a Saturday morning sometimes.

Rob Collie (00:30:22): Yeah, yeah. This is one of the things, like with the timing of having you on, it's always a little weird to have one's direct competitors on your podcast. How does it really benefit both of you? I'm so, oh incredibly sensitive to this that I think it's better to just not do it. Now, in some small degree like P3 Adaptive and Pragmatic Works, we do still compete a little bit. We offer training services. The percentage of our business that's training now is the least significant bit in the equation. But you're talking about how depressing it is to sell companies and then have to reinvent yourself. I've struggled with depression my entire life, I've never sold anything, so maybe I should stay away from selling things. Sounds awful.

Brian Knight (00:31:07): No, it's an opportunity though too. We've been in the same jobs for so long. When you have a chance to reboot that, it's a big life moment. You get people come and go from our employees all the time for you to actually have opportunities-

Rob Collie (00:31:19): No, no. You.

Brian Knight (00:31:21): ... jarring.

Rob Collie (00:31:22): That's right.

Brian Knight (00:31:22): Yeah.

Rob Collie (00:31:22): Yeah. I'm committed as the scrum saying goes right. Are you familiar with it?

Brian Knight (00:31:26): I am. I forgot that in one analogy. The scrum analogy of the pigs and the chickens and the-

Rob Collie (00:31:30): When it comes to making breakfast the chickens were involved but the pigs were committed. All right, so training. Last time you and I talked on the phone it was very clear to me that training is a major love of yours. How does one go from being the software passionate member of Pragmatic Works? What's that transition feel like and what is it about training that appeals to you so much?

Brian Knight (00:31:51): There's something that when you see that light bulb come on with something, somebody about some kind of technology. There's something you said about that home run where you get somebody to really open their eyes and see how to do something and be able to do it themselves, versus have to hire me as a consultant. It's really, really cool to enable others and to grow them to where they can become something more than what they were a week ago potentially. That's why I just love, love training. Even when we do what pseudo consulting projects, it's always about training the customers' hands are on the keyboard the entire time. So, just really, really enjoyable to get those light bulbs on.

Rob Collie (00:32:26): We try to be the same way. If the client wants to understand how this thing works, that's exciting.

Brian Knight (00:32:34): Yeah, yeah.

Rob Collie (00:32:34): Let's transfer that knowledge. Let's explain it as opposed to hoarding it as a means of holding them hostage so that when they want to change the default number format on a measure, they have to hire you again. First, that's also boring, right? It is boring.

Brian Knight (00:32:50): And our job is to work ourselves out of a job in many cases. And maybe that was always my job in the past. You want to build enough scripts to where they don't need you anymore to where you're not stuck in that mundane stuff.

Rob Collie (00:33:00): So, you write enough scripts and then one day you're just sitting in the server room playing Doom over the fastest internet available, and no one's the wiser, right? That's what the whole goal of ... That's the final boss.

Brian Knight (00:33:13): That's it. For the record, that wasn't me that installed Doom on that server.

Rob Collie (00:33:17): To let yourself go.

Brian Knight (00:33:19): That's your databases. That's all I do.

Rob Collie (00:33:21): Okay. So, Power Apps, sounds like that's become a passion. You're preaching to the choir here because I also think it's pretty exciting, but what is it about Power Apps that sort of scratches the itch for you?

Brian Knight (00:33:32): We had this app that we built 13 years ago, 15 years ago when we first started Pragmatic Work. It's called our CMS application, and it was always the app we gave our most junior developer to kind of maintain and the King Rat. It was just this big, huge thing. You pull the tail just a little bit and the whole thing falls apart. That one system controlled our commissioning, our training schedule, everything in our back office was controlled by that. What got me excited about it was when I first started doing it, my first prototype app was to replace that application. We just finished consulting, I finished [inaudible 00:34:05] as a consulting company, so I better go ahead and learn how to do this myself. We rebuilt that whole application it took us 15 years to build in about three or four days. And that is a story that I hear from customers tell over and over and over again that ability to build something really great in just a fraction of the time.

Rob Collie (00:34:24): There's some magic in there. I had a similar experience with Power Pivot in 2010 that led me to unexpectedly change course from, I'm just blogging about this thing to, oh, this is going to be the focus of my career for the indefinite future. And it was rewriting a single measure that had been implemented before by a consulting organization I'd hired. One measure, one formula that had taken two weeks to get it right in the real world originally, and I did it in 20 minutes by myself. And that's when I sat back in my chair and went, "Oh, the world's going to change." It took me a very long time to reverse engineer why it had been so much faster. It wasn't like DAX was so much more advanced than MDX.

(00:35:14): The thing was, and it again, I've had to tease this apart over the years, was that the domain knowledge of what the measure was doing, which happened to be sports, happened to be NFL rushing yards. I knew that domain and I was the one writing the formula. Whereas in the consulting case, the consultant who was awesome at the tech had no idea what American football was like. And so I had to transmit mind meld back and forth asynchronously, all of this domain knowledge, and he had to sort of project back at me the technical stuff, like what can and can't be done and blah, blah, blah. We had to do this weird Vulcan mind meld, which was very, very, very inefficient and that's what chewed up 99.9% of the time. What do you think the similar magic is that compresses 15 years of development into a few days with Power Apps? Where's the free lunch?

Brian Knight (00:36:03): I think it's democratization technology. It's the Power platform is taking these things that used to take you six months or a year to build or months to build it democratizes it. Whether it be Power BI for you or Power Apps for me, if you have that domain knowledge, you can build it yourself. And most people we train are not developers or DBAs. They're going to be analysts and project managers and people that you would never have thought in our space before. But it's awesome. It's democratizing diversity and technology. About 50% of who I train are women and you get more diverse audience than you ever would have saw when I crack an open Visual Studio in the past or when I crack an open Management Studio, it's so much more appealing to all platforms.

Rob Collie (00:36:47): That sounds very familiar. That sounds like through the same magic. Let me see if I can understand correctly. When developers would join your organization, you would assign them to this app, and regardless of how technically sound they were, perfect coders. They don't understand the domain of running the business.

Brian Knight (00:37:06): Right.

Rob Collie (00:37:06): The tremendous amount of tribal knowledge in your head for instance about how the business needs to operate. And so you have that same specification transmission problem that I had when I was doing the original MDX-based consulting project on this football thing. But then when you sit down with hands on keyboard writing a Power App, was it just you or you said we-

Brian Knight (00:37:30): Oh, that one's just me there. We got others that joined later also.

Rob Collie (00:37:34): Okay, now you're writing it yourself, but you sort of understand the business requirements implicitly?

Brian Knight (00:37:40): And you've probably been on a project before where you're like, the amount of requirements I have the right to communicate this, to give to you, to develop it are going to take me so long to write it, can I just build it myself?

Rob Collie (00:37:50): Exactly.

Brian Knight (00:37:51): Where it landed with you and myself on the app side.

Rob Collie (00:37:54): It's neat to see that it's the same dynamic. And this is Microsoft's whole strategy these days. Is kind of just unreal that you can build a technology that is accessible to the audience we're talking about, and at the same time scales in robustness all the way up to IT is okay with it.

Brian Knight (00:38:15): And you're starting to see even more democratize as they add things like chatting into it. Where you can say, "I want an app that does X, Y and Z," and it starts to build components for sorting things and filtering things for you without having to write a single line of code. And the same thing we've had that in Power BI forever with the semantic searching and those kind of things.

Rob Collie (00:38:35): It's really kind of amazing. I don't know that Microsoft can be stopped in a good way. This is not the evil empire of Microsoft, the '90s. This is the kinder, gentler version. I mean, I just think it's game over. I just don't know who can compete with them in the modern enterprise space over the next decade. I mean, who knows? One of our next topics I'm sure will be the impact of AI.

Brian Knight (00:38:59): Yeah, that's going to be interesting seeing the stuff that OpenAPI project can do. It's going to be interesting to see what does Google look like in 10 years. If you can do OpenAPI and chat with the bot and it gives you those kind of answers that are unique answers, why do I need to Google anything at that point?

Rob Collie (00:39:15): I had some friends send me last night a screenshot of them working with ChatGPT saying, "Tell me a little bit about Rob Collie." And it spits out these two paragraphs. The amount of stuff that was published on the internet that had my name in it years ago is so much heavier than today. And so it's got this moment in time kind of snapshot, doesn't know that it's P3 Adaptive now. So it's all the descriptions are Power Pivot Pro, but the second paragraph it spit out was just pure bullshit. It was believable bullshit. But it's like, "Rob Collie and Bill Jelen host the Power Pivot Power Hour on an ongoing basis," and talking about this and this. I'm like, "What? There's never been any such thing."

Brian Knight (00:40:00): There's going to be content at least, right?

Rob Collie (00:40:03): So scary. But what about the auto coding stuff? Have you used anything like a co-pilot or anything like that you tested, any of those sorts of things out?

Brian Knight (00:40:10): I'm just now starting to try those now and they are. They're pretty cool. But you have a long way to go, but it's going to be pretty neat. It makes you wonder what jobs are going to be safe for our kids' generation.

Rob Collie (00:40:21): Yeah.

Brian Knight (00:40:21): You can pitch a fast food being revolutionized by robots, but AI could definitely revolutionize our fields to where you'll be setting things up. But beyond the setup, what happens next? For adding new columns, do you need a developer? Yeah, I don't know.

Thomas LaRock (00:40:34): I just had this conversation as I brought my boy home from his first semester at college. Now, he's a music major. We somehow got on the topic of the OpenAI stuff and how it's making its own pictures and drawing. "And so these people, these artists, they feel threatened," he says, "because anybody can make art." And I'm like, "Eh, that's not really." And I try to tell him. I go, "I'm in an industry where for 25 years they've said we're just going to automate every DBA away. They've made progress and yet we're still here. There's fewer of us and we're aging out." And so I tried to tell them, I go, "The thing is you have to think of it differently." So Rob, what you just said, the second paragraph for you is bullshit. That's because right now this is nothing more than a more easier to use and fun way of using Wikipedia. Because it just takes shit from the internet, which may or may not be true, and it gives you in the natural language style and you're like, "Wow, this sounds believable," but you don't know.

(00:41:33): But yet, you can do the critical thinking and in your case you have the domain experience. So, the open AI when people are like, "Oh my God, I could use this to do all these things." But that's only true if you have the domain experience to know what's bullshit and not. So for example, you could type into OpenAI and Brian would be sitting there going, "This pop-up box says repair database, what do I do?" And you open the AI, "Absolutely press that thing and it'll fix everything." But yet Brian didn't have enough domain knowledge to know that he shouldn't be touching shit on the production server, and so he just did it. So anyway, I had this conversation and specific to music, and I tried to explain to my son, "In 10 years you think that AI is going to be replacing artists and musicians and you're ready to fight this battle?"

(00:42:22): I go, "You got to think differently. It'll be a tool in your toolbox." Because anybody can walk in type and say, "Give me a picture of a woman sitting on a beach with a dog in a nice sunset." And there it is and it'll look so great. And you think that's amazing. I go, "But the artist will be like, 'But I need the sky to be a little more surreal and I actually want it at dusk and then I need this other thing.'" And they're going to have the picture. They're going to be able to supply more of the details into that image and they're going to use AI to create a better image than just the regular person. And he's going to be able to do the same thing with music and to not be against it from the start, but to see it as a tool in your toolbox.

Brian Knight (00:43:05): Is it possible that we can have AI replaced Kanye West at least.

Rob Collie (00:43:09): Easy.

Thomas LaRock (00:43:09): I think we're there.

Rob Collie (00:43:10): Easy layup.

Thomas LaRock (00:43:10): I think we're already there.

Rob Collie (00:43:12): Yeah. Did you see Chappelle on Saturday Night Live? He's like "I'm not saying that Kanye is mentally ill, but I suspect that he might be unwell." So, I think artists should be scared. It's not because I want artists to be scared. I don't want the world to work out that way. But my assessment of all this is like, okay, so take the metaphor that we've been talking about. There was a point in time where even though I had been adjacent to the MDX story, I had worked at Microsoft, I'd been the liaison between Excel and the analysis services team. I had access to all of those people, the people who created MDX, and I still couldn't learn it. So there was this priesthood. I had visions of what this football cube could do. If we had the football cube, I knew what that would do for us, what it would do for humanity, ladies and gentlemen. So, I sort of had the story to tell, but I couldn't write it. It was trapped in my head.

(00:44:15): And then DAX comes along, suddenly now I can write that story. The audience of people on the planet that could create that football cube story expanded tremendously. The number of people who were actually good at MDX in the world, probably 10,000 people who said that they were good at it and probably a 100 who were. In the Will Hunting thing, there's a huge difference between Marco and Alberto and me in terms of what we can do in DAX. But it almost takes one of these people to know the difference. Most people can't tell the difference between me and them. I think that sort of dynamic, that democratizing dynamic is coming for the arts. I have very vague songs in my head. I'm not capable of composing. I'm pretty sure that especially three years from now, given the pace of things, I'm going to be able to sit down and provide tuning instructions to an AI to produce something that sounds really damn good. Now, I still can't sing. We've got freaking celebrities who have been doing nothing but programming EDM, like Deadmau5 and-

Brian Knight (00:45:21): Daft Punk-

Rob Collie (00:45:22): Major Laser, Daft Punk. And then there's also Diplo, right? I got dragged to a Diplo concert one time. I just laughed the whole time. Afterwards, everyone's leaving, my whole friend group is like, "Oh, that was awesome." I'm like, "Yeah, you know, haven't lived until you've seen live Diplo." The recordings just don't do it justice.

Brian Knight (00:45:50): I want to have a look at that.

Rob Collie (00:45:54): I was so cynical about it. I was trying to harsh their euphoria so hard. I'm like, "Gosh, that was the most pathetic use of our time possible." If those people can be stars, why can't there be a whole new generation of it? Chris Ray, who we've had on this show, he's doing this thing on Facebook right now where he's telling one of these AI art generators to produce him fictional cars that have never existed, and these things are coming out as photographs.

Brian Knight (00:46:22): Geez, Ali's doing that.

Rob Collie (00:46:24): I don't know if it's that or Midjourney or-

Brian Knight (00:46:26): Wow.

Rob Collie (00:46:26): He was doing this months ago and he is showing the before and after of the same things, how much better it's gotten in the last six months. And I mean, shit is scary. It is scary. I agree with you, Tom, that it needs to be a tool in the toolbox. You also have to anticipate the social change that comes with it. Which is there are people who have three out of the four skills that are required to be a talented musical composer, and this one's going to give them the fourth one.

Thomas LaRock (00:46:52): Right.

Rob Collie (00:46:52): Now, I'm still going to want to go to see an orchestra, seeing human beings with these inanimate objects in their hands suddenly make that symphony, it's just unreal. We just went this weekend, the vocalist that came out front, I mean, just unbelievable. Having machines do that will not carry the same visceral impact for the audience ever. Because there's a connection to the human that's happening there. But if it's a recording, oh, I don't know. I think for recordings, I think it's wide open.

Brian Knight (00:47:25): Like Kanye West.

Thomas LaRock (00:47:27): He might be unwell.

Rob Collie (00:47:30): We could train a Kanye bot in about five minutes, couldn't we?

Brian Knight (00:47:32): That's true. That's true. Right?

Rob Collie (00:47:33): Just spew random, narcissistic, hatred.

Brian Knight (00:47:40): Well, in other words, just be on Twitter. Okay.

Rob Collie (00:47:43): Yeah. So Power Apps, that's the same dynamic.

Brian Knight (00:47:46): It is, yeah.

Rob Collie (00:47:47): Bring it closer-

Brian Knight (00:47:47): Bring it back home.

Rob Collie (00:47:48): ... to the domain expert, yep. I am very much interested in these things like Copilot on GitHub, super, super, super advanced, auto complete.

Brian Knight (00:47:58): That's what my worry is, is all the mundane crap going to be done by a bot like that? And then the actual thinking of how to pull that stuff together is going to be the human. At least for the next 10, 15 years maybe.

Rob Collie (00:48:10): But see, I don't know. It's just we're so bad at estimating exponential growth.

Brian Knight (00:48:14): Yeah.

Thomas LaRock (00:48:15): We are.

Rob Collie (00:48:16): The self-driving car problem, is it going to be like Go? The game Go, and chess, or is it going to be like self-driving cars? So those are two very interesting examples that at one point in time they all seemed unsolvable by AI. And then chess got just solved, just straight up solved. And no human could beat the chess AI anymore. Just couldn't be done. But at that moment, no, no, no, no, no, Go. "Let me tell you folks, Go is the one that's not solvable. That game is not solvable. It's not like chess." And now the Go AI is kicking the crap out of the best Go players. Whereas self-driving cars, pun intended, the last mile has proven to be incredibly challenging. Way more challenging. Actually, that one was predicted to be solved, unlike the others where it was not solvable. This one was predicted that was going to be solved within a matter of time and they're really, really struggling still.

Brian Knight (00:49:15): They are. Speaking from my self-driving car. It's kind of scary sometimes. You have to really watch what it's doing, otherwise you could be in a heap of trouble.

Rob Collie (00:49:23): Certain problems might ultimately come down to, you need the general AI to solve it. You can't just train it based on a training set and then expect it to deal with a slightly different variation of an unknown problem. I'm very, very, very non-knowledgeable. But my best personal rationalization of why the self-driving car problem is proven to be so difficult is that it might just reduce to we need the artificial human that's capable of general intelligence to solve this. That might be that we just underestimated the problem. I don't think that, if Go can be solved is writing code or hell building a data model, building a Power BI data model, is it more like Go, or is it more like-

Brian Knight (00:50:10): It could be, right? I mean, it's at least a starting point. Build a template. Build me a template and then I can go tweak it from there. I think that's solvable.

Rob Collie (00:50:16): I do too. I mean, we've had the auto detect relationships feature for a very long time. That sucks, right? Like it-

Brian Knight (00:50:23): 50%, yeah.

Rob Collie (00:50:25): I like to tell people, but I'm still teaching classes. This thing will be right so often that it'll really let you down when it's not and you won't be able to tell that it screwed you over. You'll waste hour upon hour trying to figure out why this thing is broken before you realize it created a relationship that wasn't appropriate. So, I don't know. I certainly hope it's more like self-driving car than Go.

Thomas LaRock (00:50:48): I asked OpenAI, I said, "Write me a story about Database Performance Tuning." And in the story they basically said, "This company hires expert and the slow performance is a combination of factors. Poor indexing, outdated hardware, inefficient query execution." And I'm just like, "Go on." It came back with an answer. Honestly, it came back with the starting point. If you knew nothing and you were like, "I'm having a database issue, what should I do?" This bot would at least point you in a general direction and platform agnostic. Look at your indexes, look at inefficient queries. Look at the size of your hardware. How much percentage of the population wouldn't even know to start with those things?

Rob Collie (00:51:34): How would you train this thing? Would you train it by pairs of questions and answers and present them to human judges that don't know the domain? And so it gets trained to the point where it's providing a credible answer to the people who can't tell.

Thomas LaRock (00:51:52): Right.

Rob Collie (00:51:52): If we sat down instead and trained a specific domain into this thing where the human judges that were providing the reinforcement and the training, the judging, good answer, bad answer. Where we were saying, "You make up some bullshit about my life, Rob Collie. And I tell you that's not right." Eventually we might be able to get a ChatGPT that was able to answer questions about me. Maybe.

Thomas LaRock (00:52:15): I'm converting that into a haiku now let's see what it comes up with. Performance tuning Haiku. It's down right now. I was trying to do it myself. It was not down, but it's already at capacity right now.

Rob Collie (00:52:24): The domain of X, Y, Z, whatever it is, fill in the variable Haiku is a ... Ah, I can't get enough of it.

Thomas LaRock (00:52:31): Yes.

Rob Collie (00:52:32): Yeah. Like DAX Haiku, the filter function, iterator of Greatness.

Brian Knight (00:52:43): Or an epic poem, Thomas.

Rob Collie (00:52:44): That's right. That's right. Haiku is next level elegant.

Brian Knight (00:52:49): What NFL team do you follow, Rob?

Rob Collie (00:52:51): I'm a huge fantasy football guy, right? And so I don't care who wins the games. All that matters is that my guys acquire, that they amass statistics. If you were watching me rooting for a game and you didn't know what was going on, you'd be very, very, very confused. Because I'd be rooting for the Jaguars to move down the field, move down the field. I saw that you were at that game, by the way. I saw your Facebook.

Brian Knight (00:53:19): Oh yeah.

Rob Collie (00:53:19): You were at that game. That's going to be the one of the best moments of ever being a Jaguars fan.

Brian Knight (00:53:22): It was. There's not a whole lot of good ones, so that was a nice one.

Rob Collie (00:53:27): No, no. But there's so much sweeter that it was the Cowboys get stuffed. I mean, you either love the Cowboys or you hate them.

Thomas LaRock (00:53:32): Even cowboys fans don't love the Cowboys.

Rob Collie (00:53:35): So let's face it, it's a religion, it's a cult. Just like the average college football fan in the south. I have discovered that in the rest of the country there are college football fans who actually don't dislike the human beings that root for the other teams. And this was a real revelation to me after living in Florida. But the Cowboys lost to the Jaguars this weekend in dramatic fashion, and it was delicious.

Brian Knight (00:54:06): They handed out free towels to soak up the tears of all the Cowboy fans. It was brilliant.

Rob Collie (00:54:10): You could sell those tears. You ring them out into a vial, say, "I've got Dak Prescott's tears." So yeah, we were talking about fantasy football, of course. And so Brian, I've written a very long and comprehensive article about using Power BI to prove that you need to be in a two quarterback or Superflex league. I'll send you the link.

Brian Knight (00:54:32): Please.

Rob Collie (00:54:33): It's like this long DAX blog post, my usual DAX type of blog post. And in the end it's all just this massive troll for the people who I was playing fantasy football with at P3 who'd been stubbornly refusing to switch to the 2QB format. We've been 2QB, actually, we've been Superflex ever since. And oh, it's so good.

Brian Knight (00:54:51): Today's my first pre-con at pass. Years and years and years ago I had myself and one other person there, we're co-teaching. Our topic was about fantasy football. We're going to do data warehousing analytics on fantasy football. And we got, we build a Power BI that could go out and snag all the data from the NFL website. And then we had the defensive coordinator for the Seattle Seahawks that came in and told us their business requirements, or so the attendees thought. It was actually my salesperson, but they had no idea, of course. And nobody knows the mid or the coordinators. Nobody knows this coordinators unless you live in the city, right?

Rob Collie (00:55:27): That's right. That's right.

Brian Knight (00:55:28): Yeah, it was great. That same year we upped to Annie, were like, we had the last session at pass that year also, how are we going to keep these people attention? The last session, the last day of pass. So, myself and my co-presenter decided we would do shots every time we had a successful demo. They were fake shots, they were like apple juice. But we go, "Demo," we do a shot.

Rob Collie (00:55:49): I was in this session.

Brian Knight (00:55:50): Oh, were you?

Rob Collie (00:55:52): This is the only session I've ever been in of yours. It's my first impression of y'all.

Brian Knight (00:55:56): Oops.

Rob Collie (00:56:00): And you even had people from the audience coming up and doing shots and-

Brian Knight (00:56:02): We did.

Rob Collie (00:56:03): ... so clearly once the audience member did the shot, they knew that it wasn't alcohol.

Brian Knight (00:56:07): Yeah, well.

Rob Collie (00:56:09): They played it off really well.

Thomas LaRock (00:56:10): Let's just say, as a board member, this was a thing.

Rob Collie (00:56:13): Yeah.

Brian Knight (00:56:14): It was frowned upon. That was the end of my sessions that pass.

Rob Collie (00:56:19): Skating out of their begging for forgiveness rather than permission.

Thomas LaRock (00:56:22): People thought they were really drinking on stage. Wasn't Patrick with you as well?

Brian Knight (00:56:26): Patrick was our price patrol. He was a down as a price patrol walking around, stumbling around. He'd had a few shots himself.

Thomas LaRock (00:56:33): So they're like, "Oh my God, it was the worst thing." And I'm like, "You know it was an act, right? It wasn't real," but it didn't matter.

Rob Collie (00:56:45): I wasn't sure. I'm not remotely calibrated to this, so I had no idea. Maybe you guys were doing it. I believe that it was fake after I was told it was fake. So, the rule came down. One is not allowed to drink or fake drink in sessions from now on. Seems like there's a lot of loopholes to get around this, right? What about fake lines of coke? Who wants to come snort a line, crush the party?

Brian Knight (00:57:15): That hour session will be done in 15 minutes.

Rob Collie (00:57:17): That's right, that's right.

Brian Knight (00:57:19): Speaking a 100 words in seconds. Yeah, that be great.

Thomas LaRock (00:57:23): This is why DBAs make a lot of money. Suckers and below. It's right here.

Rob Collie (00:57:26): Yeah. DBA Performance on Meth is the title of the session, so you get out your meth pipe.

Brian Knight (00:57:34): In a ouija board, you're ready to rebuild the master database.

Rob Collie (00:57:39): Just turns out that faking anything other than drinking is probably very difficult.

Brian Knight (00:57:44): Yeah, that's poor taste, Robert. I can't believe you even recommend that. I just-

Rob Collie (00:57:47): Yeah.

Brian Knight (00:57:47): Come on, Rob.

Rob Collie (00:57:48): I'm sorry. So what's next for Brian Knight and Pragmatic Works?

Brian Knight (00:57:53): We're teaching in other areas now. We spent, we've actually doubled in this year in size. We hire a lot of people that are from the teaching profession, like people that are high school teachers and figured that if you can keep the attention of a high school student, you could probably keep the adult attention also. And then we could teach the technology, but we can't teach you to have a personality. That's always a tricky part. It's been fun.

Rob Collie (00:58:14): It's another one of those cases of these traumatized veterans that you're trying to reintegrate into civilian life and instead you take them into the tech world. Like high school teachers, they've been through the wringer.

Brian Knight (00:58:28): They have and most of them they love teaching but they hate having that one kid that won't listen. And that aspect of, it's a tough part for teachers. So, we've hired probably eight in the last four or five months that are teachers.

Rob Collie (00:58:39): That's awesome.

Brian Knight (00:58:40): One thing I've been really passionate about is making it where employees can own their own salary. They come in with a base pay, call it $50,000 and then we have a operating plan that says, "Hey, if you get this, it's $1,500 per search. If you do a class and it's live by yourself and you get X amount." And then that all equates to them owning their own salary is what the goal is. And we have a little app, go figure Power App to maintain all that. It kind of changes the game a little bit for us and gives people the autonomy we hope to achieve.

Rob Collie (00:59:11): Well, we're even more alike we keep discovering, because we do the same sort of thing with our consultants, and guess what? It's very transparently calculated and there's Power BI reports that tell you what the incentive plan is going to do for you this month based on what you've been doing.

Brian Knight (00:59:25): Share my applicator. We have a little salary simulator also where they're going to say, "Well, what if I was a Power BI person versus the Power App person and I did this, this, this?" And-

Rob Collie (00:59:33): What if?

Brian Knight (00:59:34): Yeah, exactly, yeah. It's hard that what always breaks it is you get that one person that comes in that needs, has a special condition, but largely it works for about 95% of your audience employees.

Rob Collie (00:59:46): You design these intensive schemes and then people find the ways to game them and then you have to revise the scheme. But that's part of the game. To put yourself on the same side of the table as the employees. Management is almost always adversarial, fundamentally. If you're a manager, how do you get more out of this employee while paying them the same? You're trying to optimize performance out of the employee without paying more. If you find ways that now the employee and the manager are incentivized in the same direction, now the question is, "Let's figure out how to pay you more." But it takes imagination and flexibility and also frankly, the right kind of team. Try this in the average corporate culture and it's not going to go very well. Because everyone's been institutionalized, as they say in Shawshank.

Brian Knight (01:00:31): And if you have somebody that's a junior person that doesn't want to advance, you probably don't want that person on your team either. So, if they're stuck in that same position and they have a plan in front of them on how to make a 100,000 and they're still stuck in the mud, you may not have that curious person yet that you're looking for.

Rob Collie (01:00:46): I love the career changer. Here's the business idea for you, Brian. You seem like the one to execute on this. I think the average Starbucks barista is a very, very, very under leveraged asset. Are you a Starbucks frequenter? Do you go to Starbucks, Brian?

Brian Knight (01:01:04): I do not. But that's okay, go ahead.

Rob Collie (01:01:06): All right, so first of all, imagine, just take the caricature in your head of what the average Starbucks customer is like.

Brian Knight (01:01:12): Got it.

Rob Collie (01:01:12): Every time I go into a Starbucks, Tom, by the way, when it's winter and I see all the cars that are still running out front.

Thomas LaRock (01:01:17): Yeah.

Rob Collie (01:01:18): I think if your joke about we've now entered the free car season at Starbucks-

Thomas LaRock (01:01:22): Every morning.

Rob Collie (01:01:23): ... just walk up and take any car you want.

Thomas LaRock (01:01:24): And a lot of them are high school kids. They're on their way to school so they just stop real quick. They just leave the cars running because it's cold up here.

Rob Collie (01:01:31): Leave the car idling and-

Brian Knight (01:01:33): Where you at?

Rob Collie (01:01:34): Any Starbucks in a cold climate. You live in Florida, although people in Florida are weenies about cold.

Brian Knight (01:01:40): It is 60 degrees outside, I'm wearing a Parker right now.

Rob Collie (01:01:44): Yeah, it was 19 when I went to Orangetheory this morning. These people deal with an incredibly complicated high velocity job. The number of different configurations of every drink. I mean, there's just billions and billions of possible configurations of drinks. And the people who are in there, I would say that the audience skews a little bit entitled. And there are people who are in a hurry. It's an uptempo thing. And so every little mistake gets magnified. The pace that these people are able to operate under all this complexity with all of this you sort of social pressure interaction, I think there's a similar dynamic there to the high school teacher. Starbucks' ability to identify and train people to succeed in this job at the scale. Think about the number of Starbucks baristas they're required to keep America running. We could probably start poaching these people and paying them more to do something more valuable than making sure my mocha is correct.

(01:02:42): Which is pretty valuable by the way. I can reliably get my customized mocha made to my specifications at any location in the United States. I wouldn't say it's five nines of reliability. It's maybe like one nine.

Brian Knight (01:02:56): Pretty good.

Rob Collie (01:02:56): Which is still pretty good.

Brian Knight (01:02:56): You're going to single handedly cause the downfall of America. This is Rob Collie's America folks. No baristas anymore now Rob Collie. You're going to be a vending machine at a hospital disputing this brown substance liquid out of it.

Rob Collie (01:03:10): Because all the baristas quit and became EDM wonders.

Brian Knight (01:03:14): Rob Collie's America, yep.

Rob Collie (01:03:18): Every single DJ is going to have some caffeine reference in there.

Brian Knight (01:03:25): Master Mocha, yep.

Rob Collie (01:03:26): Anyway, no Brian, it's going to be you that brings America down because you're going to be the one that finds the retraining repurposing program where the barista has basically become the backbone of American industry.

Brian Knight (01:03:37): But I wrote that down, so now I put Rob Collie on it there too. So that way I'll, whenever this hits big, I'll blame you.

Rob Collie (01:03:45): All right, that sounds good. That sounds good. You take the profit, socialize the cost to me, sounds like a good business model.

Thomas LaRock (01:03:52): For one of you.

Brian Knight (01:03:53): Sound like Elon Musk, right?

Rob Collie (01:03:54): That's right. That's right. Yep, yep. You are on your way to Bond villain hood. Which means you're either going to need to stop exercising or get really serious about exercising.

Brian Knight (01:04:06): Or lose a lot of hair.

Rob Collie (01:04:06): There's only two dimensions possible for Bond villains. Ultra Fit or Kim Jong-un. Those are the two body types. Neither of us are a good Bond villain body type yet, but we can get there.

Brian Knight (01:04:18): We can get there.

Rob Collie (01:04:19): One of them is a little easier than the other though. I think I could Kim Jong-un it, no trouble. With Orangetheory saved me 40 pounds of fat minimum. Can I do a 100? You bet I could add a 100, no problem.

Brian Knight (01:04:30): That's Rob Collie's America though. We don't want to do that.

Rob Collie (01:04:33): I'm just going to change the formula of my mocha, Brian, we're going to go from the one pump of syrup back to the four. I wanted to circle back something you said earlier. Comes up actually reasonably frequently on this show, which is you and I both had the benefit. It's also an obstacle here, but I think it's mostly benefit of, at the times we were starting our careers in this space, there wasn't a lot of competition.

Brian Knight (01:04:56): Right.

Rob Collie (01:04:57): SQL Server Central was one of a small number of places to go in the world to get SQL Server community tech content. And when I started PowerPivotPro.com, for a while it was the only place that was consistent. There were other places where you would occasionally get some content like that. I was the only place that was dedicated to it. Imagine breaking into, trying to start our career arcs that we've been on. Imagine trying to start that career arc today. I even go so far as to say that I wouldn't even necessarily get into DAX today. I'd probably be so intimidated by the expert level DAX that you run into, it's not written in the dialect that I was so drawn to. There's so few measures out there that are simple, calculates anymore.

Brian Knight (01:05:45): Every time across everybody else with the big function they may have built, right? Yeah. Yeah.

Rob Collie (01:05:48): And the optimal nature of it. Provably this is going to perform optimally, which most of the time doesn't matter. Because you've really got to get into some very interestingly, large data sets before you would even notice a 10X difference in performance. You don't even notice it when it's milliseconds either way. There's unintentional gate keeping. But in addition to that, there's just so many sources. You're coming online today as a first time DAX practitioner. There's just so much content. I definitely wouldn't be a content producer and it's iffy whether I would be a practitioner. It's really kind of scary.

Thomas LaRock (01:06:25): But why is that? Because the market's too big? When you did it, you knew it was more niche?

Rob Collie (01:06:31): Well, for me, there was this sense of adventure. We're boldly going where no one has gone before type of thing. And there wasn't anyone to measure myself against. Thank God, Marco and Alberto took a little while to switch their focus from MDX to DAX. Because I probably would have been looking at their content and going, "Oh my God, I don't understand this, so DAX isn't for me." We need more content that is unabashedly amateur. It's okay to suck, because even sucking at this stuff is amazing.

Brian Knight (01:07:01): So, where do you start nowadays then? If you are Tom's kiddo wanting to become a performance tuning expert or want to become a DAX expert or a Power Apps person, what is the right answer nowadays?

Rob Collie (01:07:12): I don't know. I don't go looking for that stuff.

Brian Knight (01:07:16): I struggle to say I have two kids in college right now. I made it through two years and I dropped out because of .com. I was working 80 to a 100 hours a week and I'm like, "That's my hard knocks there in the case." I don't know if every technology, I mean, is it more vocational basically? Is it more, "I'm going to work for free for somebody for a year and get everything I can out of it and I'm getting more current knowledge at that point." I'm not learning Fortran or something a little more arcane. What you're going to miss there is all the algorithmic kind of knowledge, discrete mathematics and those kind of things that drive future stuff. I don't have the base foundation that the college graduates might have.

Rob Collie (01:07:52): I think that's overrated, Brian. I got the whole computer science education and it just doesn't matter. None of it matters. Really don't think it's helpful at all. I came out of computer science undergrad without any passion at all for writing code. I was the only undergraduate in the graduate level graph algorithms class. Because that was yum, yum brain candy.

Brian Knight (01:08:15): Right.

Rob Collie (01:08:16): Did I ever apply graph algorithms in my life? No. Never even came close. I remember with great glee one time torpedoing an idea that was gaining traction by using O notation to explain the complexity of this algorithm. The glee wasn't about torpedoing the idea. The glee was, I was using something I learned in college. This was the one time like, look, see, four years of college and this is it. It all came down to this moment. Even that you didn't need O notation. You could have just intuited your way through it and said, "Look, as this number grows, the complexity's going to grow with the square of that number." And we didn't need the algorithms class in college to know that, it's just-

Thomas LaRock (01:09:00): So, Rob, we're going to have, to have a whole episode dedicated to this. But I will just tell you flat out, when I was production DBA. Just becoming more senior, I flat out how the guy tell me at DEV that I absolutely should know all the different sort notations and bubble sort and all, and I should be totally understanding the level of complexity and how it affects the performance in order to be a DBA. And I'm like, "That's what I need?" Because that's what I need. That's telling me in order to be a master in SQL Server, I had to learn all that XML, otherwise I can never be a master. And I was like, "Who the fuck is in charge here? Because this is some real bullshit. You know what I need to do as a DBA has nothing to do with understanding O notation and bubble sort. I get what a logarithm is, but do I really need to have that in my head when I'm looking at the query that somebody needs help with? I don't think so."

Brian Knight (01:09:51): No, you put that in your head, you're going to forget your name.

Thomas LaRock (01:09:52): Right. It's gate keeping. That's what we have to talk about a deeper level someday.

Brian Knight (01:09:57): Well, I'm kind of curious, Tom, how much did you tip the scale with your kid's college? Did you-

Thomas LaRock (01:10:02): None.

Brian Knight (01:10:02): Did you force them into it? Did you put your hand on [inaudible 01:10:05]-

Thomas LaRock (01:10:05): My daughter went to school. She had a thought of some business accounting, forensics accounting at one point, which was like serious FBI level shit, right? Money laundering. Like, "Why are you interested in this? That's different." And my son is a musician, so neither one are leaning towards anything tech related.

Brian Knight (01:10:25): What we would call tech like IT.

Thomas LaRock (01:10:27): Right.

Brian Knight (01:10:28): IT related, yeah.

Rob Collie (01:10:28): Did you tip the scale as far as them going to college at all?

Thomas LaRock (01:10:31): Yeah, I would say we always set it as an expectation that they would continue their education or they could join the military. But they had to do something. You couldn't just sit around the house. You needed to make progression to be a responsible, productive contributor to society at some level, yes.

Brian Knight (01:10:50): Yeah, I feel you. We had the same kind of challenge as well. My son's going that kind of journey right now and he is like, he's gotten job offers now and he's like, "I can drop out and make what I wanted to make when I was going into college." And I'm like, "Ah."

Thomas LaRock (01:11:03): You could.

Brian Knight (01:11:04): And there's gate keeping, like you said, there's me gate keeping down the road. It's going to prevent him from getting that full career path he may want.

Rob Collie (01:11:10): I was hyper honest with my kids about the value of college, which is like, "Meh." My ex, who's their mom. is the complete opposite. Just like, "You've got to go get those merit badges. The merit badges are you, they're your identity." You can imagine how much respect I have for that approach. There's a reason why we're no longer married, mostly because we disagree on everything. All I managed to succeed with there is installing total nihilism in one of my kids. He's like, "Nothing matters." "No, no. That wasn't the message. That wasn't."

Thomas LaRock (01:11:49): "I got it, Dad, thanks for the lesson. None of this matters."

Rob Collie (01:11:53): "Nothing matters. It doesn't matter." Just-

Thomas LaRock (01:11:56): Well, maybe you listened a little too well-

Rob Collie (01:11:57): ... back to the Xbox. Yeah, no, no. I think maybe a little, there was a sort of a filter being applied on all the things I said. So, anyone listening to this, you want to work in computers. Please do not think that a computer science degree-

Thomas LaRock (01:12:13): Absolutely not.

Rob Collie (01:12:13): ... is some critical component in your life.

Thomas LaRock (01:12:15): It is not.

Rob Collie (01:12:16): It is not. It does help certain people get certain jobs. It might open some doors, but the number one thing that a CS degree would do for you is have you hang out with other people who are really fascinated with this stuff and have you all get really deep into programming together. And you program not because your classes are making you do it, but because you like it. That environment can be replicated outside of college too. Computer science degree did not instill in me a love or an excellence in coding at all.

Brian Knight (01:12:45): Now, I think it's more important to figure out what you don't want to do before you start that four year journey than what you do want to do. You know, just want to cross stuff off forensic accounting or whatever. I really love this stuff and that's awesome.

Rob Collie (01:12:57): It's like The Cat in the Hat thing one and thing two approach. The way to find something is to first find out where it's not. And they just keep running around putting Xs on everything where they were. The sock or whatever isn't located. It's not there. It's not there.

Thomas LaRock (01:13:12): Damn, we go deep on this show.

Rob Collie (01:13:14): Yeah, deep references. But again, we're all the same age, right? Yeah. So we were all exposed to The Cat in the Hat. The average 25-year-old is Dr. Seuss' books a part of their formative experience? Maybe, I don't know.

Brian Knight (01:13:25): Probably not. Rob Collie's America again. All we get is The Cat in the Hat references and-

Rob Collie (01:13:29): AI generated Cat in the Hat books-

Thomas LaRock (01:13:32): I'm on it.

Rob Collie (01:13:32): ... written by baristas.

Thomas LaRock (01:13:37): Write me a Cat in the Hat story. Hold on.

Rob Collie (01:13:37): Oh yeah, write me a Dr. Seuss' version of the Theranos scandal.

Brian Knight (01:13:42): Okay.

Rob Collie (01:13:45): All right. Well listen, Brian, it is a pleasure, long overdue, the longest conversation you and I have ever had.

Brian Knight (01:13:50): I appreciate it.

Rob Collie (01:13:51): I've long respected your approach to things from afar.

Brian Knight (01:13:54): Well, thanks for having me. Appreciate it.

Rob Collie (01:13:56): Thanks for listening to the Raw Data by P3 Adaptive Podcast. Let the experts at P3 Adaptive help your business. Just go to P3adaptive.com. Have a data day.

Check out other popular episodes

Get in touch with a P3 team member

  • This field is hidden when viewing the form
  • This field is hidden when viewing the form
  • This field is for validation purposes and should be left unchanged.

Subscribe on your favorite platform.