02.22.22

From FP&A to Global Director of Data, w/ Khaled Chowdhury

Listen Now:

Hello friends! Today’s episode is like no other! We learn the inside secrets to creating an internal Power BI community within an organization from none other than Khaled Chowdhury, Global Director, Data & Analytics with CMC Materials. Khaled is one of the doers who took what he learned from training, rode the excitement after class, and achieved buy-in at his company by tackling a huge project using a free resource over a weekend. When the powers that be were shown his quick, agile, and cost-effective results compared to the expensive, labor-intensive method formerly used by the company, they were sold.  Khaled is a Power BI champion of the highest level after overcoming the “it’s too cheap to be a good tool” perception that is cost-assigned value.

Pulling from his extensive background in building and transforming world-class teams, leveraging technology, and driving adoption, Khaled opines that most projects fail because new technology is seen as an invasion versus a revolution. Invasions happen from outside the organization, revolutions start from within.  Perception, he says, is always key to implementation. If you are implementing a Power Platform project of your own, you can’t afford to miss this episode.

All of this and more on today’s episode!

In this episode:

Superman Turns Back Time

Dave Gainer on Raw Data by P3 Adaptive

The Curse of Knowledge

The Great Wine Label Swap Study

VLOOKUP replacement on P3 Adaptive Blog

Analyze in Excel

Mike Miskell

I’m Winston Wolf

Big Fish moment from “That Really Bad Movie”

Keanu Reeves John Wick Training video

Borg Assimilation

Rob Collie (00:00:00): Hello friends. Today's guest is Khaled Chowdhury. I have known Khaled for the better part of six years, but it's not just me. Many of us here at P3 have had a relationship with Khaled over that same time period. In fact, I think Khaled and the companies that he's worked for over that same time period is in the top three longest running professional client relationships that we've had at P3. But I think that, in terms of career transformation that is possible with Power BI and related tools, by that metric, by the career transformation metric, I think Khaled is in the top one. Maybe it's a multi-way tie, but there isn't someone that I know whose career has hit a greater inflection point than Khaled's as a result of colliding with these tools. As just one small indicator of that career change, when Khaled and I first met, he was living in Texas. For a while now, though, he's lived almost perfectly on the opposite side of the world.

Rob Collie (00:00:57): So when we recorded this, it was relatively early morning for him and it was in the evening for us. Now, most of our shows we record during the workday. Evening recording a little different. Couple of us in the Eastern time zone got a little punchy. We always have fun on these, but that evening vibe brought something a little extra. We talked about how he's the hardest working lazy person we've ever met. We talk about how Power BI is almost hurt in terms of credibility by how affordable it is. At another point, we talk about his surprising observation that in some situations there's upside in deliberately ignoring the people that you've just trained. He's a funny, intelligent, and down to earth person. And I'm proud to say that his career and our company have sort of grown together over the years. It's been a shared journey. So we're long overdue having him on the show and it just wouldn't make sense to pause any longer would it? So let's get into it.

Announcer (00:01:52): Ladies and gentlemen, may I have your attention please?

Announcer (00:01:56): This is the raw data by P3 Adaptive podcast with your host, Rob Collie. 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 (00:02:17): Welcome to the show. Khaled Chowdhury. Two syllables, three syllables. Three syllables, but I mangled the middle syllable. Is that correct?

Khaled Chowdhury (00:02:25): That's close enough. Chowdhury.

Rob Collie (00:02:26): I feel like I got it. What? No, probably not. We're going to set the record here for longest geographical distance for podcast recording. Where are you right now?

Khaled Chowdhury (00:02:38): I kept it simple. So I'm a Bangladeshi American working in Singapore, living in Malaysia. I'm kind of on the border between Malaysia and Singapore right now.

Rob Collie (00:02:47): Wow. Okay. So it's closing in on six o'clock in the evening for me as we record this, what time is it for you?

Khaled Chowdhury (00:02:53): It's closing in on 7:00 AM for me.

Rob Collie (00:02:56): Almost perfectly the other side of the world. Listen to that. Can you hear that folks? Coming at you from opposite sides of the earth.

Khaled Chowdhury (00:03:04): I'm living in the future, man.

Rob Collie (00:03:05): Yeah, I guess you are, it's tomorrow isn't it?

Khaled Chowdhury (00:03:07): It is tomorrow.

Rob Collie (00:03:09): So what is your official job title today?

Khaled Chowdhury (00:03:12): My official title is Global Director for Data Analytics. So what that means is I have end-to-end pipeline for data analytics today. From once the data is recorded, post ERP, CRM, and things like that, all the way, how the data gets presented to the users for analysis and whatnot. From the time I met you, Rob, I have gone backwards. I was a user in finance. I've gone backwards to all the way up to owning data engineering and warehousing at this point.

Rob Collie (00:03:50): Which is kind of a fascinating journey. That's one of the questions I had for you was when we met years ago you were very much a finance professional, but you were hunting for something weren't you?

Khaled Chowdhury (00:04:03): Yes.

Rob Collie (00:04:03): When you came to that class in Texas, you were on the hunt for something. What was the itch? And what year was that? Do we have any idea?

Khaled Chowdhury (00:04:11): Unlike you I don't have photographic memory. You remember everything from every movie. But this one I do remember because this was September 28th, 2016. Because, as I told you later on, that changed the trajectory of my career. I went to school for economics, came out in management, accounting, but most of my career, I grew up in FP&A, so financial planning analysis. So quote unquote, Excel people, in the finance and accounting world. In 2016, I just got frustrated of trying to do the same thing in Excel. And regardless how big or small a company is, it doesn't matter, it runs on Excel. So on one side I saw people analyzing 4k videos real time. And here I couldn't figure out what my number was last month, or last quarter, for that matter. The itch was to find something that solved the problem. I was not trying to get into data science. I was not trying to do anything fancy. I was just trying to figure out a way to make it better.

Rob Collie (00:05:12): Yeah, you weren't saying, "Hey, I really need to get into the data pipeline." You had a finance problem. What is the quarter total number? There's even a little bit of ambiguity about that at times. That's not acceptable, even though I'm sure that's how it had been for a very long time in many, many, many, many places. So you were motivated, but what led you to take my class?

Khaled Chowdhury (00:05:33): I don't think there's anything called coincidence. I was in Boston for a conference for FBN, a conference, and Microsoft sent someone in to show Power BI. This is the first version of Power BI in 2016. And even that impressed the heck out of me because anything compared to that version of Excel. And I walk up to the guy, I kind of kept him from going to the happy hour, kept him for an hour. Like, "How should I start?" And he goes, "You know what? You should get started in Excel because you already have the two tools to get started and whatnot." So I was like, "Okay." So then I go Googling and guess what? There was a training happening in Houston. At that time I was in Dallas and convinced my boss saying that, "Hey, I need to do this." So there I was in Houston and my introduction to Rob and Austin.

Rob Collie (00:06:26): And I remember also being similarly kept from happy hour by you at the end of the class. I get to the end of a seven and a half, eight hour day, basically running a one person play, who's completely drained, but you, oh man, you were just lit up. And I didn't mind talking to you. It was really striking how much energy you had for this even after the first day of class. So I do remember, I absolutely remember that interaction for sure.

Khaled Chowdhury (00:06:56): I think there was someone else on the show that talked about this. There's things that I get excited about. There's no way I can communicate why I'm excited about this. There are so many things that I started seeing potential of. It's like, why doesn't anyone else seeing this? Right? So just trying to figure that out. And then it's like, there must be something wrong if this thing is so powerful, why it's someone else not doing it? And hence comes all my barrage of questions and I had an expert on hand, so...

Rob Collie (00:07:24): It is natural, right? You see something, and I've seen this over and over and over again, people see these tools for the first time and they have this natural first reaction of, "Okay, why am I only hearing about this now?" Because the promise of it is so great. You can sort of see that. This does lead to a very natural, I think, skepticism. Like where is the catch? Where's the weak point? Does it do this? Does it break down under this situation? You keep looking for the fatal flaw and you don't find it. But even then though, you're like, "Ugh," there's a little bit of a leap that you have to take. I know who you were talking about, I think it was the conversation with Dave Gaynor that we had about how, when I first came to know him, I was so intuitive as a thinker and I still am a very intuitive thinker.

Rob Collie (00:08:15): But in order to put your thoughts out in the world and get collaboration and buy in, and you can't just transmit intuition into the world. You have to revisit your own intuition and kind of reverse engineer it and figure out what is it that's actually the promise here? Put it into really distinct English and then project that. And that is a hard thing to do. I'm still developing that skill. And Dave was the one that absolutely demanded it. That was the point where I actually had to start doing this for the first time. Instead of just telling people to trust me.

Khaled Chowdhury (00:08:47): I have definitely gotten better over time. But I think there is a term called curse of knowledge or the curse of excruciating pain that I have been through, through experience, is trying to put all the dots together for me to convey that. That is something that I still struggle to this day. I'm trying to work on it.

Rob Collie (00:09:06): Now when we first met, so we talked about you being in finance, but you were working for a chemical manufacturing company, is that correct?

Khaled Chowdhury (00:09:14): KMG Chemicals, correct.

Rob Collie (00:09:15): Yep. When you left the Houston class, you go back to the office. What happens then? I didn't get to watch that. How did you dig in?

Khaled Chowdhury (00:09:25): Before I go into that one thing I would say is I was not what you would call driven prior to that point in my career. I was good at what I did. Got the job done and taken care of. I was curious so I would poke around. I was extremely lazy. People call me efficient and whatnot, but I do not like doing things the long way. So I would sometime code for 10 minutes that's something that I could have typed in one minute because I just didn't want to type the thing. But one thing that happened at that time was I worked for Marcelino Rodriguez, the then CFO, he helped me find my passion. So I was starting to get to a point where I started becoming passionate about what I did. I was looking for my passion. And this hit at the same time is after the training, those very rare situations, I could not wait till I get to the office in the morning.

Rob Collie (00:10:23): Yeah. Can't wait to get to work.

Khaled Chowdhury (00:10:25): Yes. So I was at work at 7:00 AM. I remember those days I was at work at 7:00 AM. I worked through my weekends and at night I was like, "Okay, well I'm going to try doing and that and all that fun stuff," right? So I come back to my boss and say, "This thing is amazing. Let me try doing this and whatnot," and he goes, "You're crazy." As luck would have it we had a problem with one of our consolidations and we were choking with 300,000 records. So I was like, "You know what? Let's put this into practice." So I took one of those first Frankenstein models. Meaning that thing had way too many dimensions than it needed to have, obviously my power query was a little shaky at that point. The proof is in the pudding. So when I went and showed that they're like, "This looks too good to be true. I haven't spent a million dollar doing this. It's like, why is this working?"

Rob Collie (00:11:20): Let's stop there for a moment. You just said, we can spend a million dollars doing this and you just did it. I don't understand. How can anything cost a million dollars? I'm not savvy. I can say the word consolidation, but I've never performed one. What was the process like before and after? Even for me, I'm also sitting here going, "Oh, come on. There's no way this process could have cost that much beforehand." What's really going on there.?

Khaled Chowdhury (00:11:45): So usually for financial consolidations and planning, you're bringing in data from different ERPs for majority of the larger corporation because they don't live in one ERP. So for them, they have to bring in data from different systems and then consolidate, meaning they have to match up the, map the numbers, summarize by month, quarter a year and all that fun stuff, eliminate intercompany and more accounting jargon. Let's put it that way. However, to implement consolidation system, there are tools that delicately does that. At that time, major of the tool, even today for a company of our scale, if you spend close to... People don't think it's not a good enough tool till you spend a million or a couple of million bucks on it. So a million is actually a low number.

Rob Collie (00:12:31): Okay. There are software systems out there whose sole purpose is to try to help you consolidate across all of your ERPs?

Khaled Chowdhury (00:12:38): Yes.

Rob Collie (00:12:39): This is like a market segment. Okay. And of course, whatever the software costs, then there's the implementation that goes with it.

Khaled Chowdhury (00:12:47): The software cost is not the most expensive part. It's a couple hundred thousand dollars, but the most expensive part is the implementation cost. Right? But the best part about those is at the end of the day, those export to Excel for your reporting.

Rob Collie (00:13:03): Just like all good, expensive reporting tools. Whenever something doesn't meet your needs, export to Excel is the answer so that you can go and take it the rest of the distance. And until you actually build reports that people aren't compelled to export to Excel repeatedly, you're not there.

Khaled Chowdhury (00:13:20): Absolutely.

Rob Collie (00:13:21): Okay. So you've got this million dollar plus potential implementation on one hand. On the other hand, you've got Excel with this extra add-in installed, and you produce the numbers and you show that to them and they're just like, "Nah, this can't be right. These numbers have got to be falsified, right? This can't be real."

Khaled Chowdhury (00:13:40): I still deal with this to this day because I usually try to get to the solution that is the simplest, because I prefer using technology that can eat mud and still move at wrap speed.

Khaled Chowdhury (00:13:53): So what I call wine label problem, which I still haven't been able to fix to this day. Let me elaborate. They did wine testing. So there was wines $5, $10, $15, $150 and $1,500 wine. And it was a blind test. They could see what the label is, but they tasted the wine, and not only they were expected to say which one they prefer better, there was a MRI hooked to their head, meaning they actually were checking the brain signals to see if they enjoyed it or not. And guess what, wine people enjoyed the most is the $1,500 wine.

Rob Collie (00:14:30): And did they know is the $1,500 wine or not?

Khaled Chowdhury (00:14:35): Yes. They can see the label. So people categorically preferred the $1,500 wine. They state it and also their responses were recognized by the MRI machine, the right portion of the brain being lit up. Except they swapped the $15 and $1,500 wine.

Rob Collie (00:14:56): Oh, that is nasty, nasty. They didn't swap the $5 wine. They're like, "Man, they might catch us on that. But the $15 wine, they won't be able to tell the difference."

Khaled Chowdhury (00:15:05): Essentially what that came down to is unless someone was a expert wine taster, which majority of the people in the world are not, people assign value to the price they pay. So one of the biggest challenge we face it is like, "This is not too expensive. How can this be good?"

Rob Collie (00:15:23): I totally get it now why you call it the wine label perception problem. This thing was free at the moment, you were paying zero.

Khaled Chowdhury (00:15:29): One thing that majority of the tools in the world goes into is summarization of summarization. So summary data comes off from different systems, keep summarizing till the end of it. Right? So what you end up doing is reconciliation of reconciliations, so it's like, okay, whose number is correct? Did that summary come out right? Did we map that point right? Versus, I mean, the worst enemy I had at that point was it was still an Excel wrapper. Meaning people like, "Oh, this is just good on Excel." So at that point, I think 2016 was about the time where people love to diss on Excel. It's starting to change again. Excel is starting to become popular again, but in 2016 and timeframe, they're like the time of that time, I was like, "We've got to get out of Excel." I mean, some company did a marketing gig around that, like why they got out Excel and whatnot. So that was kind of working against me. But the beauty of it is you never summarize in, at least I don't. My models have the granular details.

Rob Collie (00:16:29): You're not exposed to the introduced error by everyone's summaries along the way. It becomes like this game of telephone where the message keeps getting passed and garbled just a little bit every time. But you start from the raw ingredients. And even though that sounds more complicated, it's actually a lot simpler.

Khaled Chowdhury (00:16:50): Yeah. So from that perspective in audit and whatnot, they like to check how you got your data, what you did and all that fun stuff, and say, "You know what? I got the data in the most raw form. I haven't done anything to it. It has every single detail that you need." So the ultimate form of audit trial is completeness and accuracy. So the number ties. You want the detail click on the plus button, or in Power BI, drill through into the detail. It will give you by transaction, which system, where it came from, what the origin transaction was, who the people was, right? So these are all the things that was not going for me.

Khaled Chowdhury (00:17:23): So the part that was going for me was my track record because I have a finance background. And then here I am talking to my CFO and CEO about trying to do this. And then later I found out through a friend of mine, my CEO was talking to the board and he was like, "I have this young kid," I guess I was a kid at the time, but had this young kid talking about... In finance, he said he could do this and whatnot. For my track record of delivering he's like, "You know what? Let him try it out." So that was the beginning of my journey. The push.

Rob Collie (00:17:56): Five years ago, you were still a child. You were a kid. But you are now a data man.

Khaled Chowdhury (00:18:03): What I usually call people is I am a accidental data professional. That's about it.

Rob Collie (00:18:07): I think almost all data professionals that are actually good at it are accidental data professionals. I don't trust anyone that set out to be a data professional.

Khaled Chowdhury (00:18:17): No, I think I'm a full scale data professional on that. And then as the thing grows, I mean, there's just so much to learn.

Rob Collie (00:18:24): I want to back up to something you said earlier and also ask Evan a question at the same time. Which is Evan, can you believe that this guy that we know was ever lazy?

Evan Rhodes (00:18:34): No.

Rob Collie (00:18:35): I'm not saying that he's lying. I believe what he tells me, but it's really hard to square that I've always known this person, this person we're talking to, I've always known him to be, when he was on my time zone, he was awake before me, he was awake after I went to bed, he was always working. He's always been so high energy, so up tempo always go, go, go. That was the most surprising thing in the whole conversation so far. Was you saying? " Yeah, I wasn't really that into it back then." And Evan, you've spent more time with Khaled than I have. I wanted to see if you had the same reaction when he said that. Have you heard that from him before?

Evan Rhodes (00:19:10): I've heard it from him, I don't believe it. It might just be a different definition of the word lazy. It might just be he wasn't passionate about it, but not passionate and being lazy aren't the same thing.

Rob Collie (00:19:24): Well, okay. I agree. And at the same time we can probably say that in a way, who wouldn't be bored by the prior wave of FP&A? It's got to be boring. It's a slog. Until these tools came around, there was no way to move at the speed that you wanted. There was no excitement. One of my big beliefs, at least about myself, and I think this is true of most people, is that emotion, people tend to think about emotion as not being all that important in the business world or whatever. And it's totally wrong, at least for me, I think it's wrong for most people in that you either have to be driven forward by fear or pulled forward by excitement. If you're lacking both of those, you're going to sit still. You're going to be relatively stagnant. You're going to be relatively not sharp.

Rob Collie (00:20:07): You're not going to be looking for ways to improve necessarily. I do know people who don't need that, who don't need either of those. They're just sort of naturally that way at all times. And having worked with people that didn't require that pull or that push, they just sort of had this internal motor that was going at all times. That was really hard for me to be around because I was clearly not one of those people. What's wrong with me? I'm never going to be successful. Not like those people. Over the years I finally found places where I was actually excited, there was something pulling me forward, and suddenly I found a completely different version of myself. So I can get it, but I can't square it with the person that I know.

Khaled Chowdhury (00:20:45): So I feel the easiest way to put it. I am the lazy person that still puts in 80 hours a week.

Rob Collie (00:20:49): Yeah, exactly. Yeah.

Khaled Chowdhury (00:20:50): Problem is I would take whatever I was doing in 80 hours before, I probably can get that done in about 25 minutes today. But there's so much more I can do. So I am a man of contradiction. So that's the other part of my contradiction that I am not really after being lazy and not doing anything, but I am just very adverse to doing things manually. I'm the one that would work two weeks to save a day, but it's a day every week for the rest of my life.

Rob Collie (00:21:20): Yeah. I don't know if this is attributed to Bill Gates, but there's a famous saw about you want a good programmer, find the laziest one. That's what you want. So I think you fit that. Bill Gates would hire you.

Evan Rhodes (00:21:32): That's why we're all accidental data professionals because we have that same mentality of do something once fine, do the same thing the second time, and just refuse to do it a third time. So we're going to figure out how to never do it that way again. And that's how you become an accidental data professional.

Rob Collie (00:21:46): Yeah. And that's the cruel, cruel joke that is traditional Excel. The thing that gets you into traditional Excel, the thing that gets you excited about it the first time you collide with it is the fact that it took some tedium out of your life. You want to change a number and see what the impact on the final answer. You don't have to run through all the numbers again on the calculator. You just type it in the one cell and see what happens. You're just like, "Oh, that's neat. Save me a lot of time." But then it immediately... Traditional Excel, when you're doing BI and reporting type of stuff, it immediately inflicts upon you this next wave of tedium. VLOOKUP enters your life. You write your first VLOOKUP and you're like, "Hot damn, that was awesome." But by your 10th VLOOKUP you're already like, " Oh my God."

Khaled Chowdhury (00:22:30): I literally burned one of my laptops with VLOOKUPs. I turned the calculation on pause. This is in 2008, I had so many VLOOKUPs on that the machine went to calculation for like 45 minutes, and after that the next time machine won't turn on. So I took it to IT and they go like, flips it and smells it, "You burned it. Literally burned it."

Rob Collie (00:22:52): Wow. You melted it. Melted it with VLOOKUP. Oh my God. That is a cool story. The same sort of fundamental problem matching up keys across tables. You can feed a data volume to the Power BI engine that is a hundred times or a thousand times the size of that dataset that melted that computer. It just links it up in an eye blink. It just doesn't take it any work at all. There's no overhead. It's amazing. So the cruel joke on the average, traditional Excel professional, is that they self-identified by getting into Excel. They self-identified as someone for whom manual work, repetition and drudgery is offensive. And then at the second tier of Excel usage, when you get into the manual consolidation across data sources, and pivot tables, and then indexing into pivot tables to try to get a combined answer because you can't handle multiple fact tables at once. That level of tedium is the worst tedium possible. And it's being inflicted on the people who can least tolerate it. This is why when people leave that world and go into Power BI, they actually become happy-

Rob Collie (00:24:03): This is why when people leave that world and go into Power BI, they actually become happier human beings. Not just more effective in their jobs, but happy. And then they start working 80 hours a week, because of all that they can do.

Khaled Chowdhury (00:24:12): One thing I would say though, I think the folks in Excel you are talking about, they're still a minority in Excel.

Rob Collie (00:24:18): Yeah.

Khaled Chowdhury (00:24:18): The amount of people that use Excel and the amount of humans in that space that actually have the data gene, is still a minority.

Rob Collie (00:24:25): We have data on this. It's at most, one out of 16.

Khaled Chowdhury (00:24:28): I usually talk about that statistics all the time. And they're like, where did you get it from? It's like, I got it from Rob Collie.

Rob Collie (00:24:36): That's right. End of conversation. And then... You don't even smoke, but then you to take your cigarette and you throw it on the ground. And you just stomp it out, and walk off. Right? That's the end of the story. Is that how it works?

Khaled Chowdhury (00:24:48): Kind of.

Rob Collie (00:24:49): Where did that cigarette come from, God? You don't even smoke.

Rob Collie (00:24:53): Yeah. It's definitely an evening recording here.

Khaled Chowdhury (00:24:58): One of the biggest difference that becomes is, one of the worst thing that you can do to yourself, being good in Excel. If anyone found out or you built that one report that everyone loves, and you have been chained to that report for the rest of your life.

Rob Collie (00:25:12): Yep.

Khaled Chowdhury (00:25:12): So, that will actually become detriment to your growth. Because you're so good at it, everyone's like, what's going to happen if this person leaves? So keep him in the job. Don't let him move. That's all he does, or she does.

Rob Collie (00:25:24): Mm-hmm (affirmative).

Khaled Chowdhury (00:25:26): One of the most liberating part of Power BI, or even in Excel. So, you don't have that problem anymore. Because whatever you build, will refresh on its own.

Rob Collie (00:25:38): That's right.

Khaled Chowdhury (00:25:38): You can build in controls and governance, and all that fun stuff. And it generates itself. So first of all, there's two problems that you solve simultaneously. The reproduction is automatic. That's where you get the first freedom.

Khaled Chowdhury (00:25:51): The second freedom is, it changes the distribution cycle. In the old days, you would have to build a hundred different reports, for a hundred different managers. Low level security, you have one report. Everyone feeding from the same thing. Meaning that, it drastically changes the possibility of error. And obviously, it just works at that point.

Rob Collie (00:26:13): Yeah. The inevitable follow up questions, don't lead to an additional weekend of work, modifying your existing spread sheet. And as you've discovered for sure, at the same time, the art of the possible. Everything we're talking about so far, is in a way, doing what we used to do. But just doing it a hell of a lot more efficiently.

Khaled Chowdhury (00:26:36): Yeah.

Rob Collie (00:26:37): Then there's also this other category, of the things that we'd never even thought that we could do. We never even conceived of the goal of having better versions, better metrics. With noise appropriately accounted for. Its so many different things. You end up going so far beyond just a more efficient version of what you had before, it's crazy.

Khaled Chowdhury (00:26:58): I think I want to extend something that you just said. What happens is, whenever people get these new tools, they try to get to all the new stuff they have never done, without eliminating what they used to do. So till you pay your debt, you can never move forward. Because someone is going to come looking for that report, and whatnot. So, you'll have to first eliminate of what you used to do. Open up that capacity, for you to move forward. You'll have to find folks in your team that has the data gene, that's willing to pay the price and automate that part. Because essentially, it becomes on top of your job. But those are the people that will go above and beyond. And be like, well, this frees me up. Because for them, they have that internal drivers like, I get free from the slavery of data drudgery.

Rob Collie (00:27:53): Yeah.

Khaled Chowdhury (00:27:54): Then you have, okay. I always wanted to do this and whatnot. These are the things that we only did in once a quarter, once a year, and whatnot. Its done instantaneously. Now, you actually have the time to ask a question. It's like, hey, should we do it differently or not? I'm not saying that copy whatever you're doing into the new thing. Worst idea possible. Because then, you just have the old thing in a new bottle. However, what I do say is that, whatever you are doing, you have to figure out a better way to do it. I don't know whether you said this in a podcast, or one of our conversation before. I think I was complaining about all this stuff that keeps breaking in Excel. And you said, for a new feature, you get one brownie point. For losing one old feature, you lose 10 brownie points.

Khaled Chowdhury (00:28:39): Hence, Excel has a lot of old stuff built into it, that keeps on breaking. There is no point of having it. It's just that fact. The same thing happens with transformation. If you are going to give people all this fancy stuff, they'll be like, hey, why is my Excel report with data? Its like, shit.

Rob Collie (00:28:55): Yeah.

Khaled Chowdhury (00:28:55): That's why both from an adoption perspective and whatnot, you have to be able to pair what used to be there before. I'll give you a simple example. I built this. This is still in 2016/2017. At that point, it was still a one man army. So I worked with the sales team, and built out this sales report. And they were like, okay. That looks cool, nice and whatnot. But hey, can you give me this Excel Pivot table, and whatnot?

Khaled Chowdhury (00:29:27): I gave them the Excel Pivot table. It's just analyzing Excel, connected to Power BI. Because, it looked like the one they had before.

Rob Collie (00:29:33): Mm-hmm (affirmative).

Khaled Chowdhury (00:29:34): It took them about three months. And one internal champion's like, wait a minute, this is a better way of doing this. I don't have to go create this chat manually. It just does it for me. But you have to give people time to change. Because one of the biggest problem the people with the data gene faces, the people who has the data gene, tends to be not as much threatened by change.

Rob Collie (00:29:58): I agree.

Khaled Chowdhury (00:29:58): However, majority of the population are very averse to change. So for them, you have to give them a pathway. So, a slower transition. So even to this day, one of the biggest features that I loved, I think, came out this year was, analyzing Excel. The Excel spreadsheet works on Power BI now.

Rob Collie (00:30:16): Yeah.

Khaled Chowdhury (00:30:16): Because believe it or not, when it comes to Ad Hoc, Excel is still the king. So we put it out there for someone who has drag drop, and get to some answers. Go right ahead. But even as we approach major of the new problems, we will always give people time to transition into the new thing. Because, people don't like being told how to do it. So I think, one thing I have learned in my own journey into leadership, you don't have to tell people the answer. That's called, knowledge bullying. You have to guide people to the answer.

Rob Collie (00:30:47): That's right.

Khaled Chowdhury (00:30:47): So, I think you probably faced it in your life as well. Majority of my success throughout my career, has been other people. Meaning, I've been very lucky. This is one thing that I'm very lucky about, is the people that I get a chance to work with. So right out of school, I worked for vice president of sales. So here, I'm a finance guy. From the get go, screwed. Because I was a finance guy, trained by sales. So, there's few things they told me at the beginning, that has taken me a decade to understand and absorb. So one of the thing, Ian [inaudible 00:31:21] taught me is like, if you want to sell an idea, make people believe it's their idea. The other one was, numbers don't matter. This is after three weeks of sleepless night, I worked to put the budget together. And Scott Pries, the vice president, goes, numbers didn't matter, Khaled. It's about the story. The numbers have to confirm with the story. Because at the end of the day, humans are not logical creatures.

Rob Collie (00:31:45): Oh, come on.

Khaled Chowdhury (00:31:46): We would like to believe we are.

Rob Collie (00:31:48): Just look around. Everyone's behaving perfectly rationally.

Khaled Chowdhury (00:31:52): Exactly.

Rob Collie (00:31:53): I mean, we're a more rational species by the day. I don't know what you're... Yeah. Okay. You're right.

Khaled Chowdhury (00:32:01): So like I said, I was like, what are you talking about, Scott? I just spent three weeks of sleepless nights, trying to figure out this budget. And you're saying, the number doesn't matter. It's like, no. I think the question is, what is the core principle? And then, whether your story's tied to that.

Rob Collie (00:32:16): Well, we heard it right here. Numbers don't matter. Podcast adjourned forever. What have we been doing, wasting our lives like this?

Khaled Chowdhury (00:32:26): Yeah. So I think, in that hunt for the story, at the end of the day, numbers are supposed to be an unbiased proxy of the real world.

Rob Collie (00:32:38): Yes. I love that.

Khaled Chowdhury (00:32:40): So from that perspective, is number replacing your gut feelings? In the real world, because you'll never have a hundred percent numbers, never. I mean, if you have a hundred percent numbers, you're too late in the game. So question is, how do we make best informed gut decisions? Is your gut decision, have some kind of validation behind it? Or are you saying, we have done this for last 20 years, and we're going to keep on doing it. Even when the numbers are completely saying, the other way around. So I think from that perspective, the way I see it is, what is the easiest way we can get the right information in the hands of the right people, at the right time? And then, people can make informed decision. I am a big believer of transparency. However, in a corporate world, have to work through the different layers and whatnot. Because at that point, other people hold you accountable.

Rob Collie (00:33:39): Yeah. That's actually... It triggers a different question for me. We talked a little bit about, you showing up like Prometheus, with this fire thing that no one had ever seen before. And everyone looks at it with skepticism, which is totally natural and understandable. But then, that initial wave of skepticism passes. That initial circle of people that you've exposed to this, are starting to go, yeah. This is good. Have you ever run into subsequently, political infighting, political back pressure? Because someone else was invested in a different tool set, and didn't like what this new flank threat was looking like. Has that happened at all?

Khaled Chowdhury (00:34:17): Yes. I was at a conference. We were talking about different things. But the takeaway was, the problems that we face are 99% same, between different organizations. The problems are pretty simple. Same and similar, and blend. However, the people we have in different organization, are a hundred percent different. Meaning, your solutions are going to be a hundred percent different. Because of a human factor to it, not necessarily because of what the problem is.

Khaled Chowdhury (00:34:45): So from that perspective, I think the easiest way I have learned to work with it is, not fight it. Because if you try to go into an argument, I am not crazy in a have to believe that I can change someone's mind.

Rob Collie (00:34:59): Mm-hmm (affirmative). Yeah.

Khaled Chowdhury (00:35:00): So from that perspective, the route that I have always taken is saying, there's a place for both of them. And let them run their course. And comes back to, proof is in the pudding.

Rob Collie (00:35:15): Yeah. I've seen this before. Intellectually, I have learned the lesson that you should never play, my software can beat up your software. That is a terrible game. It's a losing game. I think that periodically, I can in the moment, I can still get baited into these situations. Because, the intellect isn't always ruling things. We just talked about this. Humans aren't the most rational things.

Khaled Chowdhury (00:35:43): Yes.

Rob Collie (00:35:44): I've seen others do exactly the same thing. I don't know... You probably... Nah. You never got to meet Mike [inaudible 00:35:49], who sadly passed away a few years ago. I watched him operate with this stuff. And it's exactly that. No. This power pivot is never... No. It is not going to replace Cognos. No. These are separate tools, with separate sweet spots. No.

Khaled Chowdhury (00:36:04): The original Wolf. Yes.

Rob Collie (00:36:09): Winston Wolfe. Yep. The Wolf. Mike [inaudible 00:36:11]. I spent years working on him, like, come on. Even privately backstage, I'm like, come on Mike. You know and I know that Power Pivot and Power BI, they're going to replace Cognos. Even backstage, he wouldn't tell me... He would get this twinkle in his eye, like he knew he was pulling my leg. No, Rob. No. There's not... And then one day, they put him in charge of everything. And he emailed me and said, hey, I'm in charge now. So, it might be that Cognos might go away now. He held that charade, for years. But a very important principle, they can coexist.

Khaled Chowdhury (00:36:49): Absolutely. The reason is, the one thing that we do not have, or we can go back on, is time. So the longer you spend bickering about whose tool is better, nothing's happening.

Rob Collie (00:37:01): Yeah.

Khaled Chowdhury (00:37:02): And in this journey, the best time to start, was 10 years ago. The second best time to start, is right now. The longer you bicker, the farther behind you're going to fall. So from that perspective, is starting the journey more important than getting everyone to agree? Because, you'll never going to get everyone to agree.

Rob Collie (00:37:20): Yeah. Intellectually headbutting a problem, never got anyone anywhere. So, I completely agree with that. So when it was KMG, and you were consolidating across ERPs... I'm just personally fascinated by this. How many ERPs were there, roughly?

Khaled Chowdhury (00:37:37): About 10.

Rob Collie (00:37:39): 10? Okay. And the reason there were 10, did KMG at that point in its lifespan, had it come together by acquisition and consolidation?

Khaled Chowdhury (00:37:47): So, KMG is a company that focused on acquisitions. At that time we grew organically, but we also grew energetically, targeting consolidating [inaudible 00:37:58] markets, into getting better price per control, and things like that. So our regular thing was, to acquire a company or two, every year.

Rob Collie (00:38:08): Wow. 10 ERPs. Evan, have you even heard of software that is meant to help consolidate, across multi ERPs? I wasn't even aware of this market segment.

Evan Rhodes (00:38:19): No.

Rob Collie (00:38:19): No? Really? If I've never heard of it and Evan's never heard of it, it doesn't exist. Khaled's just, he's just making this up.

Evan Rhodes (00:38:28): Yeah.

Khaled Chowdhury (00:38:28): It's decently large. It's called BPM, like business performance management. Or CPM, corporate performance management. There is an entire segment of software in that.

Rob Collie (00:38:39): Oh, that's right.

Khaled Chowdhury (00:38:40): Cognos actually, falls in IDM Cognos. Hyperion from Oracle, BPC, from SAP and whatnot. One word of advice. If you're going into that field, there are better and newer software out there. So, I have strong aversion to any of those three. The old guard, what you would call it. There are newer tools that does it better simpler. Similar to what Power BI does, from data analytics perspective.

Khaled Chowdhury (00:39:07): There are ones that you can use, similar to that. Can Power BI do a lot of this? Yes. However, you're going to get into a wine label problem, and control problem. Because by default, Power BI is so strong and flexible. Meaning, you'll be like, hey, you already did this for cheap. But now, you need two people to manage this. Because, the management never goes away. So from that perspective, if you need a consolidation software, buy a consolidation software. But for analytics...

Evan Rhodes (00:39:39): Well, and then it brings it back to your politics statement, right? Because you get these massive systems with all those ERPs, can Power BI... could you put together a much more lean and effective, and efficient solution, to manage all that data? And then, face that data and create the content? Absolutely. The problem is, you have individuals. They've sunk money into it. They've vouched for it. They built it, so they have a personal affinity towards Cognos or some other massive, over-engineered architected solution. And don't want to get rid of it, just because that's where they planted their flag.

Khaled Chowdhury (00:40:15): Even if you ask me, what is ideal solution? I would always say, do not use Power BI as your system of record, if you're a public company. Use something simple. That gets the basics of the job done, for what you're trying to do. And then, add that with Power BI for analytics and reporting, and all that fun stuff that comes with it. Because now, you get best of both worlds. You have a consolidation system that managed and things like that, for a public company, for a system of record perspective. And then, you have a Power BI that is... I will say, it's lean. Because, it has couple of hundred million record in it. But, it works like it only has a couple of cells in it. So, that's the beauty of tools such as Power BI. That solves the problem, and makes it look easy. And then, walks into a wine label problem.

Rob Collie (00:41:07): There's a point in the history here, where a couple things are going on. First of all, the KMG monster is wandering the countryside, just gobbling other companies. It's acquiring one or two a year. And it's like, oh, we're the big, bad KMG. And something else happens. An even bigger company comes along, and suddenly gobbles KMG. It's like, ah, there's always a bigger fish

Khaled Chowdhury (00:41:31): That is correct.

Rob Collie (00:41:31): Said Qui-Gon, in a terrible movie. Around that same time, KMG is again, running around the countryside, eating other companies. And it doesn't notice this giant shadow sneaking up behind it, to eat it. And at that point in your life, it felt like you'd almost run your course, at KMG. You'd solved so many problems there. And things were so dialed in, that in my personal conversations with you, a little bit of maybe that boredom or what's next, was coming into your voice.

Khaled Chowdhury (00:41:59): What happened at KMG, as it was making acquisitions and whatnot. The part that was fascinating is, we make an acquisition. It takes them time to get things integrated, and whatnot. And here we have our famous sales analytics, integrated in the next week of the integration, going in live. At that point, I have automated majority of my F Paint tasks.

Rob Collie (00:42:25): Mm-hmm (affirmative).

Khaled Chowdhury (00:42:26): I have built a lot of things. And at that point in my career, I was asking the questions, what next? So I have gone and become at that point, approved RBL hybrid.

Rob Collie (00:42:39): The tweener.

Khaled Chowdhury (00:42:40): So at that point, it was like, hey, am I an F Painter person? Or am a Power BI person? So that dilemma, I'm trying to figure out. And then I realized, I didn't have to pick one. But at that point, I had gone through the transformation. And I see this huge wave coming. This technological shift and things like that, coming. I mean, now it looks like... here says, hey, Power BI is the biggest baddest tool out there. And is demanded, and whatnot. But in 2017, major of the people have not heard what Power BI is.

Rob Collie (00:43:15): Yeah.

Khaled Chowdhury (00:43:16): So at that point, I think there was a discussion between you, me or Austin. I was talking about it. And they said, what's the next step? Doing the same thing again, at a different company? It falls back in the same concept like, doing same thing. Which, I'm very adverse to. So, I was having that dilemma in my life.

Rob Collie (00:43:37): And at that moment, fate intervenes.

Khaled Chowdhury (00:43:41): Fate intervenes. But one thing... I think I was reading one of your articles, or something like that. You cannot have a revolution, with army of one. So this is something playing in the background of my head. Because at that point, I have gone through the transformation myself. I was the excited guy, and walking and whatnot. But it was Khaled doing it, not anything more than that. So fate brings the next opportunity, at that time. I thought it was like, well, there's my ticket out. I'll get my pink slip. Get some pocket money, and then go to figure out something else to do. However, as fate would have it, I met the best manager I've ever worked for, Sam Hendrickson. He convinced me saying, hey, what you have been doing before, looks pretty cool. Why don't you try to do it over here? I was like, I don't really want to be a pure Power BI developer at that point in my career. And if you just want me to manage the stuff, this is not fun enough. For me, apart from pay, the level of fun is very important in a job. So he's like, okay, what would make it fun? I was like, you know what? You can get your Power BI reports, all the stuff that I have to do before, I'll keep them running and whatnot. But you got to let me do it my way. What's your way? And I was like, I want to create a rebellion. And then he goes, are you crazy? I'm like, well, see, a revolution is only a successful rebellion.

Rob Collie (00:45:13): Yeah. Rebellions get put down. Right? They get stamped out. But, revolutions are...

Khaled Chowdhury (00:45:21): But, see the difference between revolution and rebellion is, the ones that are successful. That's the difference.

Rob Collie (00:45:25): Yeah. That's right. We were thought of ourselves as rebels all along. We were actually, revolutionaries.

Khaled Chowdhury (00:45:30): Exactly.

Rob Collie (00:45:32): Because, we're winners.

Khaled Chowdhury (00:45:36): I don't know what happened in the back office. I was like, he figured I was crazy. But he figured I was the right kind of crazy for him. He was like, why not? Let's try it. So I stayed in FP&A. But my role instead of being running FP&A or business finance or corporate finance, my focus was analytics and BI. So essentially, how to do that in finance and whatnot. But I stretched that a little bit. And that's when Evan comes into the picture, with my relationship with P3. And the thing that we did is, instead of deploying models, at that time, we focused on training. I'll go back to something that you said earlier, about the driving force.

Rob Collie (00:46:19): Mm-hmm (affirmative).

Khaled Chowdhury (00:46:19): Pushed by fear, or pulled by excitement?

Rob Collie (00:46:21): Yeah.

Khaled Chowdhury (00:46:22): So for any transformation, you need two ingredients. Either there is panic, or excitement. And for transformation, you need both. One is not good enough. Because without panic, you are not going to get the people that are in inertia. You are not going to be able to break them out of inertia. So some people, actually, most of the people, you need some form of panic, to wake them up. And then, you need this excitement to pull them forward. So from that point on, started my full-time paid experiment, in trying to do a transformation or a revolution, of how people approach data. So give a man a fish, versus teach a man how to fish.

Rob Collie (00:47:08): Yeah.

Khaled Chowdhury (00:47:08): So I took that, and extended a little bit. So, we teach people how to fish. We give them a fishing rod, a stock pond. And be there to help. So, that is how the progression happened. And as no good deed goes unpunished, the success on of that, led me to be pulled into IT, to run the entirety of data analytics. And it's been a fun journey.

Rob Collie (00:47:38): I think I got it. In that, what you had done before, when you were at KMG was, you had applied some tools. You said that over and over again, the proof is in the pudding. If you're constantly producing these results, no one can argue against it, for very long. Because, there's the results. But then, when it's Cabot right? That acquired KMG. I'm assuming Cabot was bigger than KMG? Because, it's usually a bigger company. It's very rare that the small companies acquired the bigger one.

Rob Collie (00:48:03): ... because it's usually a bigger company. It's very rare that the small companies acquire the bigger ones, but I guess it happens sometimes. Now you have a different kind of problem, a different kind of challenge, which is I can't go do all of the same things for this bigger company. First of all, you've been up that hill. It's boring. But secondly, now there's many more hills.

Rob Collie (00:48:19): And so, is a different challenge. Now you've got to spread a culture. You've got to spread an idea, a way of thinking, an approach. It's categorically different than what you did at KMG. And by the way, most people that are good at one are not very good at the other. There's like completely different skill sets. And I could see how that challenge. That was exciting. That's the kind of challenge that gets you up in the morning to put in your 80 hours.

Khaled Chowdhury (00:48:41): From that perspective, there is obviously the fun factor and then the part that comes with legacy.

Rob Collie (00:48:46): Which kind of legacy? The technology and the ways that went before? That kind of negative legacy?

Khaled Chowdhury (00:48:51): No.

Rob Collie (00:48:51): Or the, you will remember my name?

Khaled Chowdhury (00:48:54): That second kind. The second kind. The part that was funny is because see, there's files that I built in Excel in 2008 or something like that, that's still being used to this day. That's one kind of legacy. But the part that makes me excited now is that I freed myself from the slavery. And then I was the instrument to freeing other people from the slavery.

Khaled Chowdhury (00:49:25): From my legacy is not tools. Tools will change. It's the people's life you make an impact out of. I think this is the part that keeps me going today. That is what makes me happy to see. I think this is where you and I, I think I picked this up from you. I think you are very used to seeing all those people, their eyes just learning, "My life is going to be different from now on." I love seeing that part in people's life. Is like, "Wait a minute. I don't have to do that anymore?"

Rob Collie (00:49:55): Yeah. Yeah. They just, they see their future and it's different. The more we talk about this, the more I think we need to make a full feature length movie, a complete end-to-end spoof of the movie, Gladiator starring you.

Rob Collie (00:50:10): You're talking about legacy. I see you addressing your team in the mornings, "Remember, what we do in data echoes in eternity." And you're talking about freeing people from the slavery. You're in the Colosseum with the other frightened gladiators and the doors are about to open. And you just rally them and say, "No matter what comes through that door, we're going to stick together." I've got the whole movie in my head now. We're going to have a budget of $500. Let's see what we can do.

Khaled Chowdhury (00:50:37): We got to find a million somewhere in the world that probably enough money [crosstalk 00:50:41]

Rob Collie (00:50:40): You think so? I mean, just the CGI alone, where are we going to get the CGI tigers? I mean, that's going to cost at least $500.

Khaled Chowdhury (00:50:50): I will make a confession that I still, if I haven't done something technical in the last couple of months, it will start bothering me. I am someone that cannot completely go purely into running transformation and whatnot. Because at that point, the imposter syndrome tends to kick at full scale in. I still like to spend some of my time doing something technical, even though I am not the best one at it anymore on the team.

Rob Collie (00:51:20): I feel you. I completely understand this. Every now and then I just have to get in there and do something. But it has to be a real problem.

Khaled Chowdhury (00:51:28): Yes. One thing that I still try to do is to trade the tool and do something that no one thought about using it to do before.

Rob Collie (00:51:36): This sounds like me. Yeah. Keep going. Yeah. I like this.

Khaled Chowdhury (00:51:39): The latest one that I have been tinkering with is using extract data from HR, use that in Power Query to create rules of different groups of roles. Essentially, creating roles with Power Query. Definition of table with different roles based on cost center, porting business unit and whatnot, that roles get generated.

Khaled Chowdhury (00:52:03): Power Automate, picks it up from there. Starts updating Azure security groups. It runs on an interval that deploys all. We're trying to move from name deployments to role based deployments using Azure security groups, because who wants to type in the 200 names that this report needs to go to? Versus you have a Azure security group.

Khaled Chowdhury (00:52:23): One of the thing that comes up is artists like, "Hey, how are you controlling this?" Is like, "Well, obviously certain amount of people have access to it. But now we have gone from running it once a day, we run it every 15 minutes." Even if someone goes and messes with the security group, it fixes itself in every 15 minutes.

Rob Collie (00:52:39): All right. For a little while there, I felt like we were birds of a feather. You're talking about, "Hey, let's take the tools and let's make them do something that they weren't intended to. Let's subvert them." And I'm like, "Yeah, yeah, that's me. Let's do that." And then you end up doing something incredibly, incredibly practical with it. And I'm like, "Oh, you lost me." My recent project...

Khaled Chowdhury (00:53:02): Wordigami?

Rob Collie (00:53:04): Is Wordigami.

Evan Rhodes (00:53:08): I'm really glad you said it. And not me because that was fun. You mean pulling how many times people say things in a podcast?

Rob Collie (00:53:16): Yeah. And ranking their vocabulary based on rarity of words using the English language. Of course, yes. I would definitely need to size the word relative to its rarity in the word cloud. You haven't lived, folks, until you've unpivoted after breaking paragraphs on every space. And then you unpivot that whole thing, however many hundreds of columns into one column of just single words. That just feels dirty.

Evan Rhodes (00:53:45): Yeah. I mean, it felt like a bit of a superfluous activity, but I'm just trying to get one word in the Wordigami.

Rob Collie (00:53:52): Oh, superfluous. Yes. Yeah. I see. But the fact that I said it means that it won't be credited to you as a unique all time word. It will, however, be credited to this episode as a Wordigami.

Evan Rhodes (00:54:04): It felt like a cheap steal.

Rob Collie (00:54:05): Yeah. Well, I mean, what other words can we throw in here really quick?

Khaled Chowdhury (00:54:10): Going back to it. I will tend to have one of my projects that I would try to do here and there, but majority of my time is focused more on how do I get not this project, but more like this person from here to there?

Khaled Chowdhury (00:54:26): When I took over the team, the engineering team was weird. They were not looked at as a people with excitement because the tools and things like that, if you look at it from data engineering, data warehousing, that's one of the coolest topics out there. And here, I'm looking at my team is like, they are not really the most excited, whether in the company. Or people are not really like, "Hey, I want that job."

Khaled Chowdhury (00:54:53): The biggest success I have this year is now, at least I would love to believe, they might see differently outside, but they're enjoying their job. And that's a hell of a cool job to have at the same time.

Rob Collie (00:55:07): Ooh, nice.

Khaled Chowdhury (00:55:08): From that perspective, this year has been tough. The new job, lots of changes. COVID three weeks of quarantine, all kinds of fun stuff. The best part of my job this year has been seeing the change in action. That is within my team. But apart from outside of my team is all the other people. Because people don't come to me, say, "Hey, we have a solution. We have this problem. Can you help us do this or whatnot?"

Khaled Chowdhury (00:55:39): They're like, "Hey, we want to do this in Power BI. How do we do it?" I have more demand than I can support right now. It's a good problem to have. From that perspective, I have my own theories about coming to IT from non-IT perspective. Centralized analytics doesn't work. What we have gone is centralized decentralization.

Rob Collie (00:56:03): Yes.

Khaled Chowdhury (00:56:05): It's a very centralized approach to decentralized. From that perspective, data warehousing, we can get into it a little bit, but say, keep that on the side and whatnot. But our focus is not to bring all the models managed by IT because in IT, it's never possible for us to understand. Now I saying IT and us at the same sentence. But it's not possible to understand what this person in manufacturing floor versus this person in the procurement department or this person in legal needs to do with their data. It is a very domain specific knowledge.

Khaled Chowdhury (00:56:39): What we focus more on is a partnerships like, "Hey, we'll help build with you. We'll do the 90% of the work, but I need an owner and a champion and a sponsor." And then we will slowly roll back into a point where we are supporting rather than doing that. Because that way we don't have to control who has access to the data. The owner of the information decides who they give access to. This allows us to walk around a few different minefield.

Rob Collie (00:57:09): It also sets up a first level of support. People don't come straight to you in central IT with our every little problem, because they're actually going to get a better answer quicker with the local, the ambassador that you've set up there.

Khaled Chowdhury (00:57:21): We created a Power BI community for people to ask questions. And then we have our own internal Power BI champions answering questions. We don't even have to step in a lot of times. Like, "Hey, I'm trying to do this." Like, "Well, I have done this already. You can try this and whatnot if we don't get to it in time." That part is quite fun.

Rob Collie (00:57:39): Those, I'm going to call them the ambassadors, the people that you end up partnering with directly in each business unit, what are they like? I have my own theories and how I would go about finding them, what I would look for. You know your situation much better than I do, obviously. What's the reality? What are these people like? How did you find them?

Khaled Chowdhury (00:58:00): This goes back to my comment about a revolution. Majority of the IT projects fail because it's a innovation versus a revolution. Innovation happens from outside. Revolution happens from inside.

Rob Collie (00:58:13): Yeah. I'm with you.

Khaled Chowdhury (00:58:15): Most of the time I do like, "Hey, we have this cool tool and we'll show you how to do your job better than you know how to do it."

Rob Collie (00:58:19): People love that. Yeah.

Khaled Chowdhury (00:58:22): Versus the approach that we have like, "Hey, you know your job better than anyone can. What we're here to do is I might be able to give you some tools to help you do it more efficiently. Somethings that you always wanted to do, you could not."

Rob Collie (00:58:35): Sounds so simple when you say it like that. Why is this not always the way? It just takes too much sense.

Khaled Chowdhury (00:58:44): It just comes from my own failures because there's so many times I failed. By nature, I'll be like, "Hey, I think you should try this. It has worked well for others. Why don't you try doing this? And there's 90% chance, but right now, I am riding on a wave." But before the wave was created, Evan was part of this. We went around the world and ran training and the trainings were done what we would call version 1.0 of the training.

Khaled Chowdhury (00:59:13): We custom built a training using P3s material, the one that you guys usually use for training. Plus somethings, we switched a few things around. We create excitement by showing a little bit of dashboard in a day stuff. Similar to that. We built a custom model and whatnot. And then, we would go into deeper into it. We define four roles between consumer, who is just consuming information. They just need to know how to move around Power BI and whatnot. Like, "Hey, you can do all this fancy stuff and get your information."

Khaled Chowdhury (00:59:47): And I left. Meaning anyone who wants to modify the information best way it's presented, whether that's in analyzing Excel or throwing a team file and Power BI or something like that. And then, the next role goes into model. Meaning people are building the data model, the training, which is longer. And the last bit of the training is going into Power Query's legacy, who's blending the data. That was our version one of the training plan.

Khaled Chowdhury (01:00:13): After the training, we went cold turkey for three months. We were looking for people who self-initiate. We would not make the contact after the training. We'll find who amongst them will take the idea and try to do something. The quality of the modeling or something like that, didn't matter. The question is who were the self-starter would take the initiative and try to do something with it? Because motivation is something that I cannot teach.

Rob Collie (01:00:44): Yeah. It's irreplaceable. Yeah.

Khaled Chowdhury (01:00:48): From that perspective, we would look for that. Once we find that we will do our best to make that person look like a superstar in their job, those are the first champions you need. Doesn't matter their DAKS skill doesn't matter what not. Their excitement and motivation is the key part because rest of it, you can fill. They become your internal champions and they will create the peer pressure that's needed for rest of the people to follow suit. You create your scenario for success.

Rob Collie (01:01:18): That's beautiful. That's some serious Sun Tzu shit right there. I like it. Did you ever have a situation where you went the cold turkey period and no one self-initiated? Did you ever run into that problem?

Khaled Chowdhury (01:01:29): Not really but-

Rob Collie (01:01:30): And what am I saying? I mean, after being exposed to the custom design training plan, we know the hybridization of the P3 methodology with the Sun Tzu that is Khaled, how could it ever go wrong? I apologize for my impudence. Oh, impudence, that's a Wordigami. You just know it.

Khaled Chowdhury (01:01:47): I think you have a advantage of knowing all the words. I'm not a man of big words, so that's not going to work for me anyways. Taking back here one out of 16, all our trainings were over 20 people. I played the probability right. I will have at least one out of that. But my theory about those, is that if we train 150 people and out of 150, if I get three champions, that's worth the money.

Rob Collie (01:02:18): Yeah. I agree. You can't know going in who they're going to be. Your guesses are often wrong about who it's going to be. And it's beautiful. I love the unexpected catch fire. The one that you wouldn't have expected and they're just a different person. Because again, they were laying dormant. They were bored. They lacked fear, they lacked excitement. Status quo, right? And then along comes excitement. And now they're a different person. It's so cool. I love that moment.

Khaled Chowdhury (01:02:46): Some of those people ended up on my team now.

Rob Collie (01:02:49): Oh, naturally. There's a gravity there.

Khaled Chowdhury (01:02:52): What happened is the next step was version 2.0 of the training was this is where I went from one man army, when I was part of Cabot, which rebranded to CMC. I had a team, I had a small team of about three people. What my philosophy is to train the trainer. I think one of the key part, the reason I like working with P3, because you guys don't like doing the same thing again and again. And I don't like to use the consultants to do the same thing again and again.

Khaled Chowdhury (01:03:24): We have a philosophy of trying to internalize technical prowess from that perspective. As Evan helped us train around the world, we were training some people on my team so that we can run the training ourselves, which forces my team to learn.

Khaled Chowdhury (01:03:40): And the version two of the training that we found really helpful is, when someone internal from the company is involved in the training, there is a strong response. Because people feel like, "Hey, this is not an awesome consultant coming from outside and showing how this is done. Hey, someone like me that works for this company that shows he's training us." The journey from that perspective was pretty amazing.

Rob Collie (01:04:09): Well, I mean, I can imagine, when Evan's conducting the training, I mean, the man has an aura. And have you seen that mane of hair? I mean, that's big shoes to fill. You need one of your own, you need a mortal. Yes, folks. It's after 7:00 PM.

Khaled Chowdhury (01:04:28): And people look at Evan, they goes like, "Hey, is Keanu Reeves teaching us today?"

Rob Collie (01:04:31): Yeah. Keanu Reeves, international man of mystery. In his spare time, he happens to just be awesome at DAX and M and PowerBI.

Khaled Chowdhury (01:04:40): One of his alternate roles in Matrix.

Rob Collie (01:04:42): Yeah. Have you seen the training videos where Keanu Reeves is at all these different gun ranges and stuff and he is training? There's a video of him. People are yelling in his ear, "Write the formula, write it now."

Evan Rhodes (01:04:57): "It's calculate. It's calculate."

Rob Collie (01:05:00): He's running from station to stations like, "Oh, that RAT relationship, I got to solve the many to many." And then anyway, you get the idea. We're off topic a bit.

Khaled Chowdhury (01:05:09): Fun tension. But one thing I was going to say, I think we have gone from version two, now we are entering the version three of the training. Each stage of the training is targeted differently. The initial stage was targeted finding your initial champions, your ambassadors.

Khaled Chowdhury (01:05:26): The second phase of the training was to get a larger group that would become, not the earlier adapters, but whatever the second term is. Now we have training wait list that's over 200 people. Now, we have gone backwards and said, "Well, let's do it differently." Meaning we are designing the third version of the training, which is custom, entirety of the training is custom. Meaning, first of all, we're trying to group. Or the original training was more sample data based training. Now, the version three of the training is we are trying to cluster based on department and usage.

Khaled Chowdhury (01:06:09): If manufacturing department wants training, I say give us an example of what you're trying to solve. And what we are doing right now is we found that something that an ambassador did, going back in and optimizing that model, automating the pipelines and things like that. To build that model and scale that up, and we will run a long training on how to use that information and that fun stuff. That is for wide audience.

Khaled Chowdhury (01:06:38): After that, "We would be asking who amongst you still wants to do the modeling training?" We take a subset of those people into the modeling training, where they will build a model with actual live data that is representative of their situation. We are actually working in the backgrounds to bring in the data for other plants and things like that.

Rob Collie (01:07:00): Wow.

Khaled Chowdhury (01:07:01): They go through the training. Training have been two days becomes four days. They do training for two days, two days of live workshop, as people are building their own thing. At the end of this training, people walk away with functioning models because the coin has flipped. Before I was trying to find one, now there's too many. We have flipped the inverted coin. And it also saves people time because it automatically self-selects people out who doesn't really want to learn it. They just want to be able use it.

Rob Collie (01:07:31): Yeah, that's a good point. I like that. Yeah.

Khaled Chowdhury (01:07:34): We ended up with a lot of people in the training that just really wants to use it and doesn't have the skillset or willingness to learn the details of it. You have to put in your time, you have to learn.

Rob Collie (01:07:41): People have a natural resistance to being excluded from something.

Khaled Chowdhury (01:07:45): Yes.

Rob Collie (01:07:46): Of course I'm going to sign up. "Oh, I don't have Power BI. I need to go get it." And then they start to learn what's really involved. And a lot of them go, "Eh, maybe I didn't need to learn quite as much as I thought I did. I'm happy to just use the charts." And that's fine. I get that. I understand what you mean by the inversion of the funnel. Before you were trying to blast it out in order to find some people. And now you've got a wait list.

Khaled Chowdhury (01:08:13): This is what makes Power BI possible. I think what my hypothesis to call triple S, safety, sustainability, and scalability. In the old days, you needed a lot more people to scale. Now you don't. But from a scale perspective, on a user side, Power BI does something that we could not do with Excel. Even if you build an awesome model, it required strong knowledge on the user side to be able to use it. Versus in Power BI it doesn't require that. It's pictorial. It's like, "Hey, if you can use a web browser," things like that. Yes, there is understanding concept of filters and things like that. But it's something that you do when you open up your bank account or something like that. It's more intuitive.

Khaled Chowdhury (01:08:59): From that perspective, it makes it scalable from both perspective. Even if we have less number of champions or builders or authors, however you want to call it. There's many times people starting from me. It's like from analysts to BI developer. Now there's a new fancy term called analytics engineer. I mean, people love coming up with these terms.

Khaled Chowdhury (01:09:18): But at the end of the day, you need less of them, but really good, invested folks in it. And then, you have the second layer of people that is able to take on from there. One thing that has changed has made the version three possible is, as I took over the data engineering as well, now we can meet the need of the business in time. One thing that we have done this year is we have re-wired everything we have, trying to checkbox every single tool in the world that we can possibly have into tools that can meet the need of the business.

Rob Collie (01:09:59): And you mentioned earlier, the metaphor of providing each business unit with fishing tackle some fishing lessons and a stocked pond. I imagine that the stocked pond part of that metaphor is a lot easier now that you're in charge of all data engineering. The stocked pond itself is sometimes hard to construct. Are my instincts correct on that?

Khaled Chowdhury (01:10:22): Correct. That is correct. We tried to do that with Power BI data flows, but that pond is good in some use cases, but from a large enterprise use case, it does not solve the problem of data warehouse. When I was running it from finance, from a small perspective it worked really well. But when you're trying to do it from an entire corporate perspective, it does not, because it's not designed to. From the transfer mission journey, what made me excited on the BI space for someone coming from Excel, was Power BI. I do have a second tool that I have fallen in love with.

Rob Collie (01:11:00): Say it ain't so. You have two loves? You've got a love on the side? What?

Khaled Chowdhury (01:11:07): They work best with each other. The best part is they compliment each other.

Rob Collie (01:11:13): I think I know who you're talking about. It's okay if you mention them. Yeah. I like it.

Khaled Chowdhury (01:11:17): From that perspective, I think what power BI did to the BI world, the same thing is happening in the data warehouse world. Something is coming and making it simple. There are a couple of players right now in there, but in my definition there's one clear leader. I'm going to come back to the tool in a second. But the first part I'm going to say this is what you're going to love is, before all the data warehouse were write only databases.

Rob Collie (01:11:43): Yes. One of my favorite jokes.

Khaled Chowdhury (01:11:45): Yes. Guess what the single matrixes we follow for our success of data warehousing? Active users per day.

Rob Collie (01:11:53): Okay. Of the data warehouse?

Khaled Chowdhury (01:11:55): Yes.

Rob Collie (01:11:56): I don't understand the ideal number of active users of the data warehouses that I've encountered was always like, and always few as possible. We don't want them in here-

Rob Collie (01:12:03): ...that I've encountered was always like, "No, as few as possible. We don't want them in here. They're gross." One, I use the data. That's it, the person who made it.

Khaled Chowdhury (01:12:09): Apart from the backend ETN ELT stuff, no one is allowed to use a service account, including myself or any of the data engineers. Everything is using single sign on and all that fun stuff, so the only matrix that we'd really care about is if this thing is being used or not.

Rob Collie (01:12:26): I like that a lot. We're not just the place that data goes to not be lost, the dusty warehouse from the end of Raiders or the Lost Ark. That's not the data warehouse we want. It needs to be visible, usable. You can retrieve things.

Khaled Chowdhury (01:12:40): This is pure luck. The reason I say that is there was a lot of things about data warehousing. It's been around for decades, and then the new cool thing was data lakes which some people call data swamps or folders or whatever you want to call it. And then people realize, well, the data swamps cannot be scaled for analytics. You have to have some structures, so people start the warehouses on top of data lakes and all that fun stuff. So the reason I said it's luck is because I ended up with the product that's called a lake house, meaning it has data lakes, data warehouse, data marts, all of the above at the same time.

Khaled Chowdhury (01:13:17): What is important to me, forget the technology, is what it makes us possible to do. So we have all data coming into a central funnel to a data warehouse which pushes data into different data marts. And this is what you're going to call me crazy, is we actually give certain owners of the domain admin rights to their own data marts to create their own views.

Rob Collie (01:13:45): No. There's going to be anarchy. There's going to be tanks in the streets.

Khaled Chowdhury (01:13:50): There is a schema called citizen developer, meaning there are a second set of owners. They can create. They don't have admin rights to the majority of the pipelines, but they can create their own views and things like that because there is a subset of our BI champions who are also good in SQL, so if they need something, they can build their own thing. They don't have to come... because the centralized function doesn't have enough capacity to meet the need. So there's a strong centralized portion, but we're also giving people the capability to add their own data, create their own views and whatnot. And guess what? We can trace their usage. If the usage is high enough, we can put it in the central pipeline, and we can automate.

Khaled Chowdhury (01:14:28): The simple concept behind that is I have limited resources when it comes to data engineers. That game is rigged to fail. So instead of them working on every single request, they only work on requests that actually has usage.

Rob Collie (01:14:41): You know that I'm on your side in this. I just can't resist the sarcasm. I have this image of you walking into a preschool with this big box of knives and just handing out knives to all the kids. "Here guys have fun." That's sort of the typical IT fear is that you're just going to be creating tomorrow's problem and a problem that can't be undone. I don't believe that. I believe that you're on the right side of history here which is that you can't do everything, and so you might as well try to set things up to work in a way that's not going to be chaotic. Whatever you don't address is going to be met with chaos. You might as well flow with it.

Khaled Chowdhury (01:15:25): So what am I doing? There are Access databases floating around. There are people using MySQL and all kinds of fun stuff on their stuff. Now, all the stuff, instead of going underground, we can give people a place to do it right. So I'll come back to the tool now.

Khaled Chowdhury (01:15:42): The tool is Snowflake. Forgo the technical jargon. It is simple and easy. I don't have to worry about scaling down or up or any of that stuff. It's MPP with the massively parallel processing or whatever they call it. We pay by the second. It's serverless, so unlike in a normal situation where I had to worry about 8:00 AM, refresh time, everyone's going to run refresh. I got to have this much capacity and that's how much big of a machine I need. I don't need to worry about any of that stuff. People have their own separate compute engines. As they kickoff, the cluster automatically scales up and scales down in 60 seconds. If they use it for five minutes, I pay for five minutes.

Khaled Chowdhury (01:16:25): And about the flying knife, I can put a budget on every single warehouse, so if they go crazy, guess what? It will stop them. And if I have a problem that needs 512 nodes, I can bring one up for five minutes. I don't have to buy this $5 million monster for it. And my data warehouse, my DBAs or database owners, used to spend most of their time optimizing the performance, indexing, all that stuff. It is not allowed in Snowflake, so I can have those people actively looking at what people are doing. The same thing we do with BI, we're trying help people move along.

Khaled Chowdhury (01:17:02): One of the guys on my team, Jim, he did this. We used to use other warehouse. We only could bring in six months of worth of data, so they had to build tools in a way that it can only bring six months of data. Even that took two hours to load.

Rob Collie (01:17:16): What was this previous thing?

Khaled Chowdhury (01:17:17): It's Synapse.

Rob Collie (01:17:18): This is what I was getting around to is that we've seen a bit of a shift, a lot of the talent, even at the upper levels at Microsoft that has been assigned to power BI for the past however many years, there's been almost this overnight exodus, and all these people have vanished, and now they're all working on Synapse. And so I was wondering if this story was coming around to explaining why we've seen this.

Khaled Chowdhury (01:17:46): It makes perfect sense.

Rob Collie (01:17:47): I think you start... Some puzzle pieces are coming together here.

Khaled Chowdhury (01:17:50): So, here. So what used to take two, three hours to load on the user side, because they're using Jump and other statistical tools, to load the data for six months to two or three hours. Now we have taken that down to minutes, and guess what? My data engineer is not involved. Jim is on my team, but he's on the BI side of the team. He went back and reformed the way he did it, measureless table, all kind of crazy stuff that I don't understand. But long story short, they have to go by rebuild the tools because they don't have to run for six months. It's like, run it for however long you want because the machine comes online for two minutes. It takes maybe five minutes to load because the billion record, the internet cannot push that thing fast enough. The best part, you go ask for the same query the second time, it'll return from cache. You don't even pay for it if it's the second time. So we have that improvement in performance, usability and scalability while my cost went down by one third.

Rob Collie (01:18:55): Sounds like win-win.

Khaled Chowdhury (01:18:57): So same thing happened with Power BI. Power BI is fast, good, and cheap. So from that perspective, and this is the next level of value. That's why you see the focus going into Synapse. I think Synapse will catch up in five years. This is where I say that you and I don't get any money from Power BI or Snowflake.

Rob Collie (01:19:17): We're not paid shills. We're just simple people trying to make our way in the dataverse.

Khaled Chowdhury (01:19:25): Yes, and what we love is creating value for whoever we are working with. So for you, that's a different definition of customer than me, but I'm still working with my customers to create value, and I'm going to use the best tool that's available to me. And from that perspective today, it's Power BI and Snowflake. Maybe it will be something different because at the end of the day, the right people at the data gene cares more about solving the problem than the packaging.

Rob Collie (01:19:51): I agree. If a tool comes along that's better than Power BI, we'll use it. I don't expect see such a tool anytime soon. I know what goes into making such a thing, and it's just really hard to envision, but who knows? I don't have to predict the future. We're reactive. We're agile. And one of the things that is really, really, really good about today's Microsoft is that everyone wants to be the everything vendor. Microsoft would love a world where everything ran on their software, but at least this particular version, this iteration of Microsoft that we're experiencing now, they're aware of the existence of other software platforms. They're aware of the existence of other tools. They want to play nice with those.

Khaled Chowdhury (01:20:34): Guess what my Snowflake runs on?

Rob Collie (01:20:36): Oh, here it comes.

Khaled Chowdhury (01:20:37): Azure.

Rob Collie (01:20:37): Wow. My head just exploded. But wait-

Khaled Chowdhury (01:20:41): My software runs on Azure.

Rob Collie (01:20:41): But wait, it's Azure Synapse. How... Oh man.

Khaled Chowdhury (01:20:48): So think about this. In today's world, from data warehouse, the shift has gone from platform as a service. I was in love with Azure SQL DB, even though I didn't know shit about Azure SQL before, but now I'm in love with Snowflake because now, even the part that I had to do before with Azure SQL DB, you don't. Snowflake just does it for you. So Snowflake is a [inaudible 01:21:10] SAS warehouse, meaning it completely takes the indexing, all the stuff in the backend. It's completely in the backend. The software does it rather than a human doing it.

Rob Collie (01:21:20): I'm starting to wonder if we should be moving some of our marketing analytics and trying that out and just... Because why leave it alone? It's working right now. We should probably just go retool the whole backend because it sounds cool. Maybe we should get out of Azure SQL Data Warehouse with all this stuff and go, "no".

Rob Collie (01:21:38): So wait, can snowflake run on AWS?

Khaled Chowdhury (01:21:41): So when you open an account with Snowflake, you go to Snowflake. It doesn't even require credit card. It will give you a 30 day trial with $400 in credits. I challenge you to be able to spend the $400 in credits.

Rob Collie (01:21:53): Enter code Khaled for 20% off.

Khaled Chowdhury (01:21:58): And there's no need for code Khaled. And when you're picking it, you get to pick. It's like, "Hey, which country do you want this to your data to be in? Do you want this in Azure, AWS or JCP?"

Rob Collie (01:22:09): That's flexibility there.

Khaled Chowdhury (01:22:11): So from that perspective, it's a good tool. There are other players that's coming up. So the other ones that are really good is Google BigQuery. Amazon Redshift is now trying to catch up. And then from both Amazon Redshift and Synapse serverless. They are trying to catch up to Google's BigQuery and Snowflake at this point.

Rob Collie (01:22:31): Wow, I'm going to have to get my hands in on this at some point. This sounds fun. Maybe we should, at some point in the future, do another show, and all we talk about is the Snowflake journey.

Khaled Chowdhury (01:22:41): I have spent this entire year trying to prove Snowflake's hype wrong, and I have failed.

Rob Collie (01:22:47): All right. Same thing as Power BI. Pick a tool, say, "Nah, no way." And then just use it to death. And to go, "Well, now I can't find any weaknesses." I have a theory that your number one takeaway from our working session in Fort Lauderdale, I think you came in there saying in a way to yourself, "How much better is Rob than me?" And you came away going, "Not very much." At some points in that meeting though, we did a lot of actually valuable of things. We did some interesting reports and some interesting explorations. At some point during that two or three days we were down there together, I started to get the feeling that you were actually testing me. I was being evaluated. I was being measured.

Khaled Chowdhury (01:23:32): So let me put it differently. People think I think outside of the box, but people don't understand that I have a disability. I do not see the box. So my problem is I-

Rob Collie (01:23:44): There is no box.

Khaled Chowdhury (01:23:46): Yeah, so it's like, "What is this box everyone keeps talking about?"

Rob Collie (01:23:49): So it's just this thing I keep walking through, and I get it in my teeth.

Khaled Chowdhury (01:23:53): One thing that I read is don't order from the menu. Most of the things that I have come across is... Even with P3. So what Evan would justify as "Hey, here is the standard curriculum, and here comes Khaled's customization." I love all this, but can you do this for me?

Evan Rhodes (01:24:11): I believe you've done that before even knowing what the proposal was, was, "Hey, we've got a great idea."

Evan Rhodes (01:24:17): "Great. I'm going to change it a little bit."

Evan Rhodes (01:24:19): You don't even know what the thing is, but he's still going to change it.

Khaled Chowdhury (01:24:26): Coming back to it, I think this is my contribution to Wordigami. I have pronoia. Instead of paranoia, I have pronoia, meaning I have a firm belief that I have come across people that continues to help me be successful, and for some reason, they go around the world and helping me become more successful than I was. And I don't get to take any credit for it. One thing that I learned very early in my life, that I am not smart. So it took you to get to Microsoft, but I realized this when I was in seventh grade. And then the second thing I learned, you are the sum average of the people around you. So what might have looked like that I was questioning you, trying to judge your capability, it was just me hanging out with you to learn from you. There are a lot of things I do and say, if you take out the voice tones and things like that, people be like, "Hey, that's Rob talking." What I love to do is I try to find people smarter than me, and one thing I have learned over time is smartness is not only technical smartness, it's emotional smartness and integrity as well. So I do not like hanging out with people with less integrity because, guess what? That tends to affect me as well. It was my pleasure. The longer I can hang out with you, the smarter I'll get. So yes I might have gone in, you might have felt like, "Well, wait a minute. I thought this guy was nowhere, and now by the time you're leaving, he looks close enough." I was just learning from you.

Rob Collie (01:26:06): It was fun. I was learning things about Power BI that I didn't know from you while we were there. And I was like, "All right, this probably isn't what he expected." I'll ask the other question.

Rob Collie (01:26:20): You were talking earlier about, you need there to be some panic. I'm thinking to myself, "Oh, so do you go cause some?" What happens if you need panic and there isn't any? Do you have to incite some panic?

Khaled Chowdhury (01:26:32): I think God takes care of that for you. I think in the world we live in, I think there is no shortage of panic.

Rob Collie (01:26:39): No doubt.

Khaled Chowdhury (01:26:39): Whether it's new automation taking your job away, to COVID shutting us down, to work 4.0, working from home, from KMG acquisition into CMC. Guess what? CMC is being acquired by Entegris in the next six to 12 months. So the world provides an ample amount of panic. You just need to learn how to leverage it.

Rob Collie (01:27:01): So you're of the Rahm Emanuel school. Never waste a crisis. I get it. Okay. I'm with you.

Khaled Chowdhury (01:27:06): You can call it crisis, or you can call it making the best lemonade.

Rob Collie (01:27:10): That's right. So yeah, there's another acquisition coming. If you were starting to get bored, that's actually how we should know. Whenever you start to get bored is when we should buy stock in your company. That's the signal from the outside that there's a takeover coming.

Khaled Chowdhury (01:27:25): Yes. You would've made about 40% even before the deal closed.

Rob Collie (01:27:29): There you go. When the first acquisition happened, your benevolent virus was recognized and sort of weaponized and leveraged. You've moved upstream. Maybe it happens again. Eventually when all the companies merge into one, when we reach the corporate singularity when all the companies are one company, you'll be running data for everybody.

Khaled Chowdhury (01:27:51): I don't know. I don't know. And to be honest, I have so little control on it. So I don't really care too much about it.

Rob Collie (01:27:58): No, you can't control things. You say you're not smart, but you are. And you're definitely wise. I've become a big believer in the value of wisdom as I've grown older. Wisdom beats smart every time over and over and over again.

Khaled Chowdhury (01:28:14): The only part of the wise part that I can agree to is that instead of working on other people, I try to work on myself because I know that I have no control on others. So same thing with this case of new acquisition and whatnot. The best thing I can do is to be better than I was yesterday, to create better value than I did yesterday. So that's what I'll do. And then whether the opportunity keeps me with the new company or takes me somewhere, I'm, regardless, excited about it because one thing that I didn't really get a chance to talk about is I've taken the learnings from you and couple of other folks into building pretty simple pillars of my transformation at this point. We have gone from reporting to data modeling because the concept of reporting is stagnant because you create too many data model to support every single report versus a single well built report can eliminate hundreds of different reports, so that is our first pillar.

Khaled Chowdhury (01:29:18): The second pillar, which you heard me talk about today is revolution or whatever you want to call it. I have been crazy enough to try to embrace the community. That is all the people that has got into this drudgery, try to embrace them rather than the concept of centralized decentralization. But the part that a lot of people call me crazy is I actually have gone backwards into the ERP world, tried to create partnerships, meaning the things that we were using on Oracle Discoverers and things like that, we are bringing that into Snowflake. We are actually giving the keys of the ownership to the people who owns Oracle. Hence, there is no need for my team to make sure this matches the source system because the source system lacks in giving the answer. Now I am empowering not only the community from a user perspective, even from a developer perspective, we are trying to do that.

Khaled Chowdhury (01:30:15): Last, but not the least, you need an agile platform. If you think about it this way, at the end of the day, there's some core problems. We need a EL tool or the extraction or ETL or whatever you want to call it. We need a tool for that. We need tool for warehouse. We need a tool for BI. The way at the pace of our change, every five years, there's going to be a new player in the market. That is the best, and we need to change every five years. So from that perspective, we are trying to build in a way when, say, Synapse catches up in five years, if they become the next best thing, we'll do a simple swab between Snowflake and Synapse if that comes to that. But at the end of the day, our priority and things are to creating value, whatever tool helps us do that. And that's what I love working with P3, because it's not about the tool. It's not about the model. It's about what creates value for me. And that's why I have worked with you guys for the longest time.

Rob Collie (01:31:17): Five years, right?

Khaled Chowdhury (01:31:19): Yes, well since that training, I've worked with it, and I've grown up with P3. That's what I would say. And as I have walked around the world, found some other people doing amazing things in Power BI, for some reason, they all attended training by Rob Collie.

Rob Collie (01:31:35): Oh, that's awesome. That can't be true. I haven't done any trainings. There's so many people out there that are doing amazing things now that I have no idea who I am.

Khaled Chowdhury (01:31:43): That's true. However, the part is in Power BI world, I'm a OG at this point. [crosstalk 01:31:49] So people that have reached the senior manager director levels, this is when you were running training.

Rob Collie (01:31:58): That makes sense. Well, that's really gratifying, and yeah, we've grown up with you. Our growth and your growth have been very parallel. It's a good sign.

Rob Collie (01:32:07): Hey, well, listen, I really appreciate you taking a big chunk of your morning. This is some of the smartest stuff I've ever heard from a not smart person, so-called not smart person.

Khaled Chowdhury (01:32:17): It's just accumulation of knowledge from different smart people.

Rob Collie (01:32:21): It's the Borg. It's the Star Trek Borg. Just absorb and assimilate the best parts of all the people you run into over the course of your life, and you can't help but improve. As long as you don't believe yourself to be complete. As long as you don't believe that, "Oh, I'm so awesome." If you have some humility about things, which is also known as realism, of course you're going to be learning. We've each stolen from each other. I'm going to steal some of the things that you said on this podcast. I've plugged them into the Borg hive that is me. So I really appreciate it. It's been great to catch up with you. The time zones have made it difficult, but this is totally worth it. I'm glad we did this.

Speaker 1 (01:32:57): 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.

Subscribe to the podcast
  • This field is for validation purposes and should be left unchanged.

Other Episodes

Copy link
Powered by Social Snap