Why AI Projects Fail for Finance Teams and How Power BI and Clean Data Fix It with Khaled Chowdhury

In this episode of Future Finance, host Paul Barnhurst speaks with data expert Khaled Chowdhury, a renowned leader in the data and AI space. Khaled shares his journey from finance to data analytics, offering insights into the evolving role of AI in business decisions, the importance of data clarity, and how tools like Power BI and Fabric are transforming the data landscape. Khaled’s approach to data clarity in the age of AI brings a fresh perspective to the conversation on how companies can leverage data to make better, faster decisions.

Khaled Chowdhury is a data expert and founder of Data Crafters, helping businesses, especially in finance, build clear data strategies in the age of AI. With a background in FP&A, Khaled’s unique journey spans from finance director to certified data professional, with expertise in tools like Power BI and Fabric. He’s been recognized globally for his contributions to the data world, empowering organizations to make data-driven decisions. Khaled’s experience bridges finance, technology, and business leadership.

In this episode, you will discover:

• Khaled’s journey from finance to data analytics and AI

• The importance of data clarity and the common mistakes companies make

• Why AI and data-driven decisions require a strong foundation

• How AI can both enhance and challenge business decision-making

• Practical advice on how companies should prioritize AI investments and data management

Khaled Chowdhury highlights the importance of building a solid data foundation before diving into AI and advanced tools like Power BI and Fabric. His journey from finance to data leadership demonstrates how clarity, smart data strategies, and a focus on outcomes can transform decision-making.

Join hosts Glenn and Paul as they unravel the complexities of AI in finance:

Follow Khaled:
LinkedIn:https://www.linkedin.com/in/khaledchowdhury/

Follow Glenn:
LinkedIn: https://www.linkedin.com/in/gbhopperiii

Follow Paul:
LinkedIn -  https://www.linkedin.com/in/thefpandaguy

Follow QFlow.AI:
Website - https://bit.ly/4fYK9vY

Future Finance is sponsored by QFlow.ai, the strategic finance platform solving the toughest part of planning and analysis: B2B revenue. Align sales, marketing, and finance, speed up decision-making, and lock in accountability with QFlow.ai.

Stay tuned for a deeper understanding of how AI is shaping the future of finance and what it means for businesses and individuals alike.

In Today’s Episode:
[02:12] – Khaled’s Journey: From Finance to Data
[06:10] – Defining Data Clarity
[10:15] – The AI Paradox
[14:42] – The Power of Power BI and Fabric
[18:30] – The Data Maturity Curve
[22:05] – How Companies Should Invest in AI
[25:03] – The Future of Data and AI
[29:12] – The AI Chatbot Question
[34:25] – Wrapping Up: Final Thoughts from Khaled


Full Show Transcript: 

[00:01:33] Host: Paul Barnhurst: Welcome to another episode of Future Finance. I'm Paul Barnhurst, the FP&A guy, and unfortunately this week I'm flying solo as Glenn Hopper is in California recording some more LinkedIn learning courses, which we can't wait to come out. He's done 6 or 8 now. I know he would love this interview because I have a fabulous guest today. We want to tackle talking a little bit about, you know, power BI data data fabric. We'll talk about some things around how our data expert thinks about AI. So I want to welcome our show to the guests. This is a good personal friend of mine, Khaled Chowdhury. Khaled, welcome.

[00:02:12] Guest: Khaled Chowdhury: Thank you. It's a pleasure to be here.

[00:02:15] Host: Paul Barnhurst: Very excited to have you. I think I even have a few of your stickers just around the corner here. I should have put on your shirt for today, but then they might have thought I was advertising, so I decided not to. Let me give just a brief background and he can add to it.

[00:02:27] Guest: Khaled Chowdhury: Maybe your first question is like why did you get a data guy in a finance podcast?

[00:02:31] Host: Paul Barnhurst: I know we have all kinds of people since we talk a lot about AI and technology here, but yes, a lot of people wonder about that. So he is the data guy. In fact, he's been recognized as one of the biggest data people in the world by different organizations. And he started his career in finance. He worked in FP&A for several years. He was one of the very first people very early on, I think 2017, 2018 to implement power BI to use for your planning, which is becoming much bigger today. He understands finance. He understands excel. He's a real expert there, but he's taken that and taken off in the data world. Started his own business today. Data crafters help companies, particularly finance, figure out their data strategy and grow their data. So where I want to start is I mentioned you started your career in finance. Obviously you moved to data analytics. Do you want to add a little bit, maybe give us the 2 or 3 minute kind of journey you took to get where you're at?

[00:03:28] Guest: Khaled Chowdhury: Let me kind of start by saying that some people are born and they know they want to be XYZ. I am not one of them. So from my perspective, my approach in life most of the time is to find things that I enjoy doing that creates value. But whatever I do, I try to probably go 200% at it. So that probably could set up the background because guess what? I even go to school for finance. I went to school for economics. My undergrad is in economics. I have a minor in finance accounting. And then as you are kind of mentioning, my background is that I can probably create three different resumes. That would be for three different jobs because from one perspective, I'm a certified financial planning analyst professional, I'm a certified management accountant and then as a finance director and all that fun stuff. So if I want to go back into finance, that's kind of one side. If I wanted to be the data side, I probably have way too many Microsoft certifications, right? Being with Fabric and Power BI, probably since its inception right now, on the other side, if I wanted to be a CXO on the data side, I'm a certified chief Data officer and all that fun stuff, right? But people might say that I have changed careers, but actually never have. I still remember the very first time I started, I was a management accountant for a company that's trying to implement the concept from Europe.

[00:04:58] Guest: Khaled Chowdhury: Right. And we were just trying to help businesses get the information they need to build and make better decisions. Right. So coming back from that perspective, think about how I started in FP&A. I just went really, really deep on at this point, the way my journey into my current thing is that I say I'm an accidental data professional, right? So that happened from the fact that in 2016, I saw people analyzing 4K videos real time. I can figure out what my number was for sales for a customer three months from before. And here I am writing the 10-K and whatnot. I'm like, ah, hopefully I don't go to jail for this. And it was telling me that in three years when they finish deploying SAP, they will get me the numbers. And this was the exact same thing he told me three years before. So I was told the same thing in 2014. A14. And at that point I was like, you know what? I think this is Kennedy who said, don't ask what the country can do for you. Ask what you can do for the country. So I was like, you know what? I'm not going to ask it to do anything for me. I'm going to go figure this out. I was fortunate I came across Kavya in 2016, started my journey there, and to be honest, from 2016 to 2018, my job, I was still a finance director, but I picked up a hobby and a second job into the bi world.

[00:06:20] Guest: Khaled Chowdhury: It came down to my project, which had 25% of the budget of the IT, and they had only one European scope, I had 11. I killed that project in less than three months and obviously my budget was so small I could not get consultants to do it. I figured out why not have the consultants guide me how to do it. And that's kind of how I went from not knowing what BI stood for in 2016 to the place we are today. So that's kind of my journey into data. And then the next part happened is, uh, an acquisition when I made it to CMC. They really liked what I did. And they're like, do you want to do this? I'm like, yeah, I don't want to be a developer. I'm a finance director. Why would I want to be a developer? At that point? They're like, well, you can do whatever you want. I'm like, okay, that sounds interesting. So the first thing I did is I got a budget not only for consultants, but I got a budget for training. So one of the biggest fallacies of organization is we implement millions of dollars of tools, spend very little time or money in letting people know how to use it.

[00:07:34] Guest: Khaled Chowdhury: Essentially, I worked for one of my CTOs, Jeff, and he used to say, like, we'll get Lamborghinis and get a monkey to drive it. So that was awfully, not awfully. I think I was a victim of my success on that one. Right. We got really successful. And then they're like, well, you know what? Why don't you just move to it? So I became a chief data officer at that point. That journey continued. And after the last acquisition, I wanted to try something different. I think, in FP&A, we always have the thing about getting a seat at the table. The coach put me in, like, what if I get to decide what gets on? So I guess, uh, from that extreme perspective, I am on the table now. So the simple concept came from the journey of Bart, of data crafters. The fact is, the projects that took 5 to $10 million in 20 1516, by the time I was CTO, cost about 2 to 3. The difference now is in today's world, that 2 to $3 million projects now cost $200 to $300,000, meaning the tools have changed significantly in the favor of people on how to leverage them. And that's what we do at Data Crafters now, helping people get the work done, because the concept of having a plan for data and AI is not restricted to fortune 500.

[00:09:03] Host: Paul Barnhurst: Thank you for that background. And it's amazing how far we've come. Like, you know, 15 million to a couple hundred thousand. And obviously lots of companies struggle with AI with data. We all know data can be messy. It's not uncommon to hear that. So I thought it was really interesting. On your LinkedIn profile you say you help or you say, I help CXOs build data clarity in the age of AI from chaos to clarity. But how would you define data clarity? What does it mean if a company has data clarity? What do you think about that?

[00:09:38] Guest: Khaled Chowdhury: Clarity comes from two perspectives. One is from how you are making decisions. Right? And essentially the chaos comes from the fact that, if I go talk to a company, say, $100 to $500 million in range. That's how much money do you spend for your data team? That would be like not much. Right? Then if I ask the question is how many of your finance team and IT team spends their time answering questions where they are doing the human busts, there's no analysis involved and things like that. And then when you look at that thing it is massive. So for a company that's in the half $1 billion range, there was probably a couple of million dollars of inefficiencies. And the second part that you notice is the fact that on the exact leadership perspective, when you try to run a company is when you try to decide, how often do we have that? Like, hey, wait a minute, my number says, we did 20 million last year. Why does yours say 17.5, so it sounds funny, but it's still like I saw this in 22,008 and I still see this today. Right. So people seem to spend more time trying to figure out whose number is correct rather than actually making decisions. Right. So from that perspective, so from a chaos perspective, it's about simplifying and leveraging your talent in the best possible way. So that efficiency goes up, how the information gets from the source to the hands of the people. The second part I usually work with from their perspective is that they want to be data driven. Now, what does it mean to be data driven? So one of the part that is a common misnomer is that data driven companies make all the decisions based on data, nothing else.

[00:11:41] Guest: Khaled Chowdhury: Unfortunately, if you're doing that, you have missed the boat because data is always behind looking. So the part, the clarity that a lot of leaders have to understand is the question is how informed is your gut? Meaning are you making? Because we work with incomplete information, things change and are not finalized. But do you understand the exposure of your decision? Have you known that the decision you're trying to make, we have tried that ten times and failed miserably ten times. What are you doing differently to make that a success? So from that perspective, that's kind of the two sides of it, right? One is more technical in terms of how we leverage your data assets the most prudent way. So that way the total cost of ownership is lower. And on the other side is what I usually tell them is like because everybody is really interested about AI, fabric and all this fancy stuff, and it's like, I can get you all this information. What decision would you make differently? And if the answer is I'm not going to do it, I'm going to keep doing the same thing. I'm like, don't waste your money here, right? Because unlike your tools of being able to produce more and things like that, data doesn't have any of that. Like it's kind of like having hands. An extra pair of hands creates more widgets. An extra pair of legs gets you to more destinations. The question is, when I give you an eye and a brain, are you making the right widgets, going to the right destinations? And it's as simple as that. I think what we say to be our mission is that we get the right data at the hands of the right people, at the right time, so they can make better decisions.

[00:13:28] Host: Paul Barnhurst: I love that the better decisions I still go back to very often. Stu West, the former CFO of Grammarly, said finances' goal is to help the business make better, faster decisions, and data is a key part of that, right? The idea behind data is it should help drive better decisions. If you're not using it for decisions, if it's not driving better decisions, don't invest a lot of your CapEx in your data, if it's going to help you make better decisions, spend the money there. It's amazing how often it boils down to a really simple concept, but it can become really complex as you get into it, you know? But often you just ground yourself and go back to those first principles and ask yourself what's important here and let that guide you. You'll generally be right and make the right decisions. All right. Everybody loves data these days. You talk. You know, we want to start with data clarity in the age of AI. But lots of people are struggling with the AI side, right. Everybody wants all the insights. We want them now. So you had mentioned you gave a presentation recently. We were talking a little bit about this beforehand, the AI paradox. Let's take a minute and talk about that. Can you tell our audience what the AI paradox is?

[00:14:42] Guest: Khaled Chowdhury: Let me kind of maybe step back. Maybe I'll go ahead and write this out as an article. So there was this concept of proactivity. Paradox, right. So think about it. I'm not talking about ten years ago. I'm talking about 100 years ago when we went from steam engine to electric engine. Sorry, not engine electricity. So when in the old days in all the mills and stuff like that, the majority of the power came from steam engines. And then we have this revolutionary thing called electricity with lines and whatnot. And we are supposed to get productivity. People go ahead and put electricity in and then they're like, wait a minute, there's no productivity. What the heck is going on? The simple paradox here is if you do the thing that you did because of the old system's limitations, you are not going to get the productivity. Meaning the lines didn't have to be straight anymore with electricity. The line doesn't. Pressure doesn't have to build up. So the line efficiently designed for electricity is different from the steam engine. The same thing is happening today in the realm of productivity in today's world. So the easiest way to see the biggest difference right now is. Let me ask you a question. How often do you use an AI tool that makes you productive?

[00:16:08] Host: Paul Barnhurst: I used one this morning, I'm sure daily.

[00:16:10] Guest: Khaled Chowdhury: So for me, it's kind of almost like a half hourly thing that I use tools that make me more productive, whether it is copilot. To summarize my thing, I have a quick idea for a blog post or a thing. I don't want to go look up the link. It goes, fetches it out, cleans it up, and I'm a horrible typer, right? I love using whispers when I'm talking to write things out. Now you see that on one end. On the opposite end, if you look at the MIT study, 95% of your AI projects fail.

[00:16:42] Host: Paul Barnhurst: We talked about that in a previous episode. I saw that study. Gartner said 40% of all projects. Right. The reality depends on who you ask. Get different numbers, but there's a high failure rate. That's the bottom line in corporate AI projects.

[00:16:55] Guest: Khaled Chowdhury: Correct. So now the question becomes why is that the case? It's the part that you realize is when you're doing it. For one, there is no concept of master data, data cleanliness and all this stuff. Right? And then if you look at a lot of the things that you can do with cloud today, cloud or GPT desktop and stuff like that, it's mind blowing. The problem is for an enterprise scenario, if you are a mom and pop shop and you want to kind of upload a spreadsheet and do something for your clients, you can. But guess what's going to happen for any company to upload their data into GPT, right? So you're not going to get away with that. The AI paradox on the professional side becomes, is that the company is trying to figure out is like, I spend this much money and why is this not working? So a couple of things are true. First of all, we have not. We are not ready to do the process differently with AI. But even before we get there, you see some of the fascinating things that the folks are doing with cloud or ChatGPT and whatnot.

[00:18:07] Guest: Khaled Chowdhury: That is great and amazing. But what about data security and the governance, right? So from that perspective, you cannot do that. The other problem people misunderstand, and this is I will try not to go too technical on this one. However, I'm pretty sure you talk about this all the time as probabilistic versus deterministic. Meaning if I open up a hardware report or excel file, the calculations are exactly the same. If I select the same filters, it's going to show me the exact same number whether I do it or you do it. I'll do the exact same thing based on your access. Two. Same person. So think about our first agent. Went live at one of our deployments, right? This is where we have a data agent, right? Data agents are slightly different from knowledge based agents. Data agents are slightly different from knowledge based agents. I'll talk about it a little bit later. But now from that perspective, as we're testing, two of us, same access, asking the same question because we copy pasted it, it was giving us different answers.

[00:19:20] Host: Paul Barnhurst: You hit on something there, the different answers. That's a real challenge.

[00:19:24] Guest: Khaled Chowdhury: So the question comes because it is probabilistically understanding and asking the question to the engine, and how it interprets the question itself means the engine might give, even though the engine is a deterministic engine, the data agents behind is deterministic. So you're writing semantic models or SQL queries against it. Those are deterministic. If you run the same SQL query, it's going to give you the same answer every single time. But how that SQL query or how that Dax is generated by the agent is still probabilistic. So that is a very fine grain. Now what I say about it today is very, very good compared to what I saw last year. What I assume is in the next 3 to 6 months, these are essentially production ready meaning it is not going to answer. So for example, if I ask the question, hey, what's my sales? I didn't ask for my life to date sales. I either asked it because remember, humans don't give context when they talk. So we can train agents to be like what I used to do five six years ago. I would interpret this as either for this year or this month. You're asking the question. I will automatically exclude inter companies, right? I would exclude essentially the context in things.

[00:20:49] Guest: Khaled Chowdhury: Right. And when you answer the question, what you figure out is if you keep asking questions back to the person, did you mean this. Did you mean that everybody gets annoyed. Rather, it can be trained to answer the question, well, your sales for this month is $25 million, excluding intercompany and whatnot. Would you want something different? It is getting frighteningly good. However, the key is how clean is your data? What I talk about is how clean your data is. And then for the realm of fabric and power BI, maybe I'll kind of cover that together. But this is one of the things that's a very high position to use. But it works very well. I mean, to be honest, except if your data models are a lipstick on a pig, Copilot doesn't look at the lipstick. It goes straight to the pig. Meaning if your data model design is bad, it'll give you really, really bad answers. So from that perspective, in the realm of AI, how we design power BI fabric anywhere, any data model, whether you're using any other tool, the data model and things like that gets even more important because you cannot hide this behind lipstick.

[00:22:04] Host: Paul Barnhurst: Ever feel like you go to market teams and finance speak different languages? This misalignment is a breeding ground for failure in pairing the predictive power of forecasts and delaying decisions that drive efficient growth. It's not for lack of trying, but getting all the data in one place doesn't mean you've gotten everyone on the same page. Meet QFlow.ai, the strategic finance platform purpose-built to solve the toughest part of planning and analysis of B2B revenue. Q flow quickly integrates key data from your go-to-market stack and accounting platform, then handles all the data prep and normalization. Under the hood, it automatically assembles your go-to-market stats, makes segmented scenario planning a breeze, and closes the planning loop. Create air-tight alignment, improve decision latency, and ensure accountability across the team.

So that's a concern. You got deterministic but we're getting there. We're figuring all that out. But the other issue you mentioned is data. So, you know, lots of people want to know where to start using AI and power BI, and particularly where companies should invest their money. You know, should they be investing a lot in AI? Should be. They focus on learning power BI. So what's the advice you give companies? I'm sure you get asked all the time, hey, I need AI and you're probably like, let me walk you through a process here. Let's talk about it. So how do you think of guiding companies with, you know, we've had fabric in the last couple of years. We see power BI, we see AI use cases all over in data. And everybody is saying, hey, AI, look at me, I'm awesome. So what's kind of the advice you give to companies on this journey?

[00:24:16] Guest: Khaled Chowdhury: The part they need to understand, especially with AI which is founded in your data, you cannot skip the maturity curve. So what I mean by that, let me break this down. Say you are in a country where you only got rickshaws, those tuk tuks or whatever you want, right? Now. All of a sudden now the restrictions are limited, you can go buy a Tesla. I mean, just going from tuk tuk to a Tesla. As long as you have roads and stuff like that, it doesn't mean that you actually have to go buy diesel engine and whatnot.

[00:24:49] Host: Paul Barnhurst: Those you have to go through all the steps, the inventions to get to the Tesla. You can go from the bike or the rickshaw or whatever it is straight to the newest car.

[00:25:00] Guest: Khaled Chowdhury: So the easiest example I can give you is that if you were using SQL servers 20 years ago, right. And then we use Azure SQL, then we use this and whatnot. There's no point of going through that evolution. You can bypass, for example, a tool, whether it's snowflake, Databricks or even fabric, which we are specialized in. You got a Lakehouse Warehouse Kql database. Everything built in doesn't mean that you have to figure out what was done 20 years ago, that evolution is not needed when it comes to the actual material, which is your data. You have to start with something descriptive. Can you get your reports from this thing? So what we call the Gartner's famous one we would call descriptive, diagnostic, predictive and prescriptive. Right. This is a ladder we cannot skip.

[00:25:59] Host: Paul Barnhurst: So you're telling me I can't get to predict without doing the others?

[00:26:03] Guest: Khaled Chowdhury: Well, you can, except it might be telling you to take the right turn right on the cliff. Go right for it. So I think that's kind of the challenge when the AI is grounded in your data and your data is bad, you are going to produce, for lack of a better word shit at an unimaginable pace and intensity that could end your business. And this is kind of where I think I can give you an idea, is that people are looking at automation right in the realm of AI. If you use the ChatGPT or cloud in agent mode, it can open a website, go do this and whatnot. And they are very different from what we call deterministic automation that we used to have, which means it's looking for specific things, right? Rather, what it is doing today is that it looks at the entire image, codes the entire image, every subsecond, and tries to make a decision. Guess what happens if you have bad data and bad web and what it ends up doing? So the part that we say today is that if you fail that deterministic automation before, you should go try that again, because there is a portion of the things that deterministic failed. And if you inject the probabilistic cover of the communication, the deterministic process always fails because it fails at human to machine communication. And this is kind of what's best highlighted in a data agent where it is doing something deterministic.

[00:27:42] Guest: Khaled Chowdhury: What it's doing is interpreting a question, passing it on and back and forth. Right. And that is one of the ones that works really well. However you try to do this on your power BI model today, you can go test this out. Just go turn on like a fabric F2 at 300 bucks a month. You can test it on your copilot like in your power BI today. Now you ask you what is my sales by territory? You have 20 measures that say sales. You have 12 columns that say territory. Which one is the right one? So I think that's kind of the idea that for a lot of people that we are doing now, and it's essentially getting different. What I mean by different is that we used to say, hey, you need to put foundations in For say a year and then get benefit out of it. It is not that long. However, if you go straight into an AI project without putting foundations in, I don't want to be part of it. I'll just come out and say it. But what I suggest people do is what we call a hybrid tap and plumbing. So as you build the plumbing takes time and investment. If you solely focus on plumbing, you get nothing out of the tap. You do not get to iterate. So from that perspective, go do the thing that has been best practice for the last 20 years.

[00:29:08] Guest: Khaled Chowdhury: Centralize the data. But when you try to do that, I walk into people. They're like, oh, I have 200,000 tables from across 30 ERPs. I'm going to bring all of them in and then try to get value out of it. I'm like, I spend millions of dollars without knowing if there's anything to use over there or not. So what I suggest is to take a centralized approach to bring your data together, but have a focus. Let's figure out sales, financials, HR, AP, something as you kind of make the investment into your estate versus that. And you can get those pretty quick, pretty fast. If somebody tells you to implement a foundation takes three months, six months, which is to take me like I spent $100,000 just running a prototype at my when I was a CTO, if that. And we are even working on coding to make it an hour. We spend two days, three days having a discussion on what you want. Understand your problem. Once that's codified, the code deploys, right? So that essentially is changing. So from that perspective my suggestion is to understand where you want to go based on that, set the right foundations with the goals that are achievable and at the same time, don't get distracted by the most shiny tool. Try to figure out how you can leverage the probabilistic with deterministic to take you the furthest possible.

[00:30:41] Host: Paul Barnhurst: I think this is a great place. I'm going to summarize here and then move on to. We have a fun little section I'm going to surprise you with, but we'll just take a minute or two here. And so I really like it and I think everybody should listen to this. This is an AI podcast. We have a data expert here. We didn't talk a lot about power BI. We didn't spend a lot of time talking about fabric. What he mentioned there is to understand your problem, build the foundations, then add in all those things that can help and you can do the plumbing. Well, you get some benefit from the cap. Don't start huge, start small. Don't spend 30 million bringing everything together before you know you're going to get value out of it. So it comes back so often to the fundamentals. Don't just buy into the hype. And I will tell you if you need someone to help you with the data finance problem, call it fabulous. He's helped me. We've had a lot of great discussions. So now here's the fun section. And I'm just going to ask you one question today. Usually you get one from me and one from Glenn, but you're going to get one. So we pump your information. I like to call it information about AI. I use Claude today. We ask it to come up with 25 kinds of personal little quirky questions we can ask you. Okay, here's how it works. You can pick a number between 1 and 25. Or the random number generator can. So see we're allowing you to be part of the process if you choose so human in the loop or all AI, which one do you want to go for?

[00:32:03] Guest: Khaled Chowdhury: I'll pick a number.

[00:32:04] Host: Paul Barnhurst: All right. Go ahead. Give me a number between 1 and.

[00:32:07] Guest: Khaled Chowdhury: 2511.

[00:32:07] Host: Paul Barnhurst: 11. And I haven't even reviewed these questions. I've reviewed a few, not all of them. So it could be interesting. Wow. Okay, I have to say, Claude really took the Houston thing to an extreme in the questions. I noticed that. So it says, Houston, we have data. What's the most Houston thing about running a data company from Houston? Why? It used Houston three times in that question. Again, we'll see the oddity of AI. So I'm going to read this one more time. And then we're going to wrap this question. It says Houston we have data. What's the most Houston thing about running a data company from Houston?

[00:32:46] Guest: Khaled Chowdhury: One of the great parts I think about Houston is it's too big.

[00:32:50] Host: Paul Barnhurst: It is a big city.

[00:32:51] Guest: Khaled Chowdhury: And part of running the data company. What it allows us to do is that we have a small but a very thriving community, both in power BI and fabric. So from that perspective, we have some amazing people like Andy, Helen, and we have one of the powerful BI megastars, if you want to think about it, Patrick LeBlanc. He is in Houston now. So key part I would because I usually end up speaking or participating in different areas and whatnot. The data community committee in Houston is very much like the city. It is. It is down to earth, no bullshit. Very open to learning and sharing. So from that perspective, being able to do that in a city that's very cosmopolitan, open and open to sharing, it's helpful, right? So from that perspective, the user group. So if you're in Houston want to come find out about fabric do find me or the media thing. But when you talk about Houston we talk about food.

[00:33:58] Host: Paul Barnhurst: Oh there are a number of questions in here about food too. So we didn't get one of those. You picked number 11. So I appreciate you trying to answer that. I realize that was a unique question. And that's what we mean. Ai sometimes gives us what we want, sometimes gives us a little bit of strangeness, but we like to bring that to the show. So Khaled, thank you so much for joining me. I loved the conversation. I'm really excited to share this with our audience. I know Glenn will enjoy hearing it, as he didn't get to join us today and keep sharing your message and helping people understand how they can get better with their data. So thanks for joining me.

[00:34:33] Guest: Khaled Chowdhury: Thank you for the opportunity. And like I said, I think we are moving from First Horizon to Fifth Horizon so fast. I'm looking forward to having a chat probably early next year, when I can get through some of this probabilistic nightmare that I'm dealing with.

[00:34:49] Host: Paul Barnhurst: We'll look forward to that next chat. So thanks again.

[00:34:53] Guest: Khaled Chowdhury: All right.

[00:34:53] Guest: Khaled Chowdhury: Thank you very much.

[00:34:55] Host: Paul Barnhurst: Thanks for listening to the Future Finance Show. And thanks to our sponsor, QFlow.ai. If you enjoyed this episode, please leave a rating and review on your podcast platform of choice, and may your robot overlords be with you.