What the Excel workflow of the future will look like with Rosie AI CEO and Excel expert Dennis Jiang

In this episode of Future Finance, hosts Glenn Hopper and Paul Barnhurst welcome Dennis Jiang, co-founder and CEO of Rosie. Dennis shares how Rosie helps professionals navigate complex spreadsheets, debug formulas, and optimize financial models. The conversation explores how AI is transforming spreadsheet use, streamlining tedious tasks, and empowering users to focus on strategic insights. Dennis also explains the unique challenges of integrating AI into financial workflows and how Rosie is changing the landscape of Excel usage for power users.

Dennis Jiang is the co-founder and CEO of Rosie, an AI-powered assistant that helps professionals understand, build, and analyze spreadsheets with ease. Dennis started his career as a consultant at Bain & Company, where he became an Excel whiz, so much so that he later worked as a freelance Excel consultant for high-profile clients like Apple. Having seen firsthand how much his clients and colleagues rely on spreadsheets yet struggle with them, Dennis founded Rosie to leverage agentic AI to finally solve that problem and help professionals focus on insights rather than wrestling with formulas.

In this episode, you will discover:

• How Rosie’s AI assistant redefines how professionals use Excel.

• The role of AI in auditing and debugging complex financial models.

• How Rosie saves time for finance teams by automating tedious tasks.

• The difference between Rosie and tools like Microsoft Copilot for advanced users.

• Why Dennis believes that spreadsheets will remain central to finance, even as AI technology evolves

Dennis shared his journey from consulting to founding Rosie, offering an inspiring look at how AI is transforming Excel modeling and financial workflows. His insights into automating tedious tasks, debugging complex models, and enhancing financial analysis provide invaluable guidance for professionals looking to leverage AI to improve productivity and decision-making.

Follow Dennis:
LinkedIn - https://www.linkedin.com/in/dennis-jiang-0387923/
Website - https://www.askrosie.ai/?utm_source=podcast&utm_campaign=futurefinance1
Special Offer for Our Listeners: Enjoy 1 free month of Rosie Premium (a $20 value) using the code FUTUREFINANCE at checkout.

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

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:
[03:04] - Why Rosie Was Created
[05:22] - Microsoft’s AI Integration Struggles
[06:45] - Rosie vs. Microsoft Copilot
[10:05] - Managing AI Hallucinations with Rosie
[12:55] - Rosie’s First Fan Club Website
[17:08] - The Future of Entry-Level Roles in Finance
[20:50] - Modeling with Punch Cards
[26:23] - The Future of Financial Analyst Roles
[33:05] - The Aha Moment for Rosie
[36:21] - Wrapping Up and Free Access to Rosie

Full Show Transcript: 

[00:01:50] Host 2: Glenn Hopper: Welcome to Future Finance. I am Glenn Hopper, along with my esteemed colleague, the FP&A Guy himself, Mr. Paul Barnhurst. Our guest today is Dennis Jiang. Dennis is the co-founder and CEO of Rosie, an AI expert for Excel modeling that helps users understand, build, and analyze spreadsheets with ease. Dennis started his career as a consultant at Bain and Company, where he was an Excel wiz, so much so that he later worked as a freelance Excel consultant for clients like Apple. Having seen firsthand how much his clients and colleagues rely on spreadsheets yet struggle with them, Dennis founded Rosie to leverage a genetic AI to finally solve that problem and help professionals focus on insights rather than wrestling with formulas. Dennis, welcome to the show.

[00:02:39] Guest: Dennis Jiang: Thanks for having me. It's a pleasure.

[00:02:41] Host 1: Paul Barnhurst: Yeah, we're really excited to have you join us. I know I have to see a demo of the product. You and I got to chat a few weeks ago now. I think it was, and I know Glenn will check it out as well. We're big enthusiasts of all the tools out there, so Glenn and I get to talk a lot about them. I want to start with this question because you and I talked a little bit about this prior to the episode, but I'd love for our audience to get to hear this. Yep. Why did you create Rosie Eye? And how does that differ from Microsoft Copilot? You know what? What's the purpose of it?

[00:03:11] Guest: Dennis Jiang: Yeah, it's a great question. As Glenn mentioned, Rosie is an AI expert for Excel modeling. Many people think of her as, you know, Chatrapati for Excel. And for those more technically minded, cursors for Excel, for those familiar, you know. But what that means for everybody is that you get this Excel expert, you know, that finally kind of like helps you kind of with your Excel and the way that, you know, oftentimes everyone kind of like probably you and everyone kind of in the audience has this Excel expert in their life. You know, sometimes you are an Excel expert. You know, like probably the folks on this podcast are, you know, but most people kind of struggle with Excel and have something in their life that they go to for, you know, help. And so, you know, but these Excel experts, you know, are rare. They are costly. And, you know, they take vacation, they have jobs, you know. And so, you know, imagine if you could have that person there by your side 24 over seven who's always there to help. So that's what Rosie is built for primarily because I've had so much experience, kind of like, you know, myself and kind of with my colleagues who, you know, have this challenge. I know what I want to do in Excel, but I don't know how to do it, and I don't feel confident about it. And so, you know, Rosie is that kind of person and kind of like we have, you know, just talking to a customer a couple days ago who was just telling me, I was like, hey, I love Rosie.

[00:04:18] Guest: Dennis Jiang: Like I talk to her like she's my close friend. And that's, you know, exactly how we designed her, which is, you know, phenomenal. You got some like, you know, social implications of that kind of relationship but world of Excel, it's strictly a positive. You know, one example just kind of points to is we were analyzing a financial model that was published by the US government recently for a $3 billion loan related to a government program, very, you know, like the standard model, you know, looks great. But we asked Rosie, it's like, hey, like, can you find any mistakes in this model? And she's like, yeah, like it looks like you're calculating interest expense incorrectly. And that's actually who she identified as. Like I said, like a $200 million error, you know, and it was like, hey, can you fix it for me? And it's like, hey, like, you know, this $200 million error, you know, for a $3 billion loan. Like, you know, that's a life or death for. You know, for a company. You know, if that loan doesn't come through because someone is a copy-paste error, you know, very subtle. And so, you know, imagine if you could always have that extra pair of eyes, you know, this Excel expert pairing with you all the time. That's what Rosie is. And kind of like people are having a phenomenal time kind of using her and working with her on their their spreadsheet problems.

[00:05:16] Host 2: Glenn Hopper: Wow. Paul, I'm going off script already. I gotta say this.

[00:05:19] Host 1: Paul Barnhurst: I was going to go off for a second here too. So you go first.

[00:05:22] Host 2: Glenn Hopper: So what you just said, I mean, Microsoft invested 15 billion in OpenAI out of the gates. You know, they've done the aqua hires of the other companies. You know, this is where they want to go. And yet when, you know, I do training around the world on AI usage and how to use it in your workflow, and it's I'm bullish long term, I guess that Microsoft will get there. And I'm now like making sure they're never going to be a sponsor of this podcast. Paul. But I'm but I mean, it's kind of crazy that we haven't seen better integration with Copilot in Excel right now. And I'm. You're too young to remember Clippy, but I'm sure you're familiar with Clippy. Yeah.

[00:06:01] Guest: Dennis Jiang: I'm older than I look, so yes, I remember. Yeah, yeah.

[00:06:06] Host 1: Paul Barnhurst: When do you have to give me PTSD? Today? By picking up Clippy.

[00:06:10] Host 2: Glenn Hopper: For the audience members who may be too young to actually remember Clippy. That was Microsoft. And actually, even if they weren't around, the memes are still around. But it's the infamous office assistant that, you know, didn't quit. It was a little early for AI, and I really mean, I think, you know, Microsoft is trying to get back to that. They've been stewing over Clippy since they pulled it from the office all those years ago. And I feel like now, I mean, what you're saying with Rosie, it's like Clippy revenge, finally delivering on that original promise of this AI assistant that actually helps with spreadsheets. And I wonder, you know, I know you're not inside Microsoft, but, I mean, why do you think Microsoft hasn't gotten there yet? And I guess how you're different. What you're building. Different, obviously, from Clippy. But from what? What's going on with Microsoft, and why haven't we seen more integration from them?

[00:07:02] Guest: Dennis Jiang: Yeah, yeah, it's a great question. Yeah. I mean, the way that I, we think about it and kind of have, have heard from the industry is, you know, the way that Microsoft is building with copilot, you know, like, absolutely no shade You know, there are a billion people who use spreadsheets on a regular basis, you know, between Google Sheets and Excel. You know, most of those folks are novices. Most people are not good at Excel. And so Microsoft's building fundamentally could be, you know, you know, like, not not the negative kind of aspiration, but like, you know, the positive aspiration of someone who helps you with Excel. What we're building with Rosie is something for the power user, the CFO, the accountant, like the professional. Those folks are not using Clippy. You know, like, you know, you know, in the way that, you know, like most the majority of the billion people use spreadsheets, kind of like do kind of like need help with with kind of more, more basic things, you know, like, so like right now copilot is very good at helping with pivot tables. Just phenomenal. A lot of people struggle with pivot tables. Even experts struggle with pivot tables, you know, but it's not the kind of thing that the CFO is going to rely on every day in the way that, you know, like what we're building is really kind of like people. People will, you know, you know, people are making billion-dollar decisions, you know, off of what they're doing in Excel. And so that's the way that we think about it, and kind of like going deeper, kind of there, is kind of where we're going. So I think that's kind of how we think about it differently. And, you know, I think there's a lot of space actually, you know, for serving like this, this more kind of advanced user.

[00:08:24] Host 1: Paul Barnhurst: So kind of how to think of that as copilot is really more for the broad audience, kind of everybody, which involves a lot of people. They're using Excel sparingly or on a basic level. And Rosie, AI is really to help the more advanced user, somebody using a lot more, going deeper with it. That makes sense to me. And, you know, we mentioned clipping. I tried to get the guy who created Clippy on our podcast, but he told me he had already told the story to everybody. There was nothing more. So we brought you on as our replacement. Just kidding. So we could at least feel like we covered Clippy in an episode. So that's why we had to bring Clippy in here.

[00:09:04] Guest: Dennis Jiang: Yeah. I mean, actually, one of the ways to think about it is,you know, like, Paul, you're wearing your Excel World Championship kind of jersey. We have this aspiration I mentioned before the show started that, you know, we want Rosie to, kind of exceed the kind of human performance at the Excel World Championship and those kinds of challenges, this year. So kind of like we're on our path to, to do so. Yeah. I can't tell the future, but I don't imagine Clippy is ever going to achieve that level. And I don't think that's the aspiration. You know, like, you know, your average Excel user does not need to like, be the Excel world champion or rely on that. But like for the expert Excel user, you really do value, you know, having somebody who's at that level. And that's what we aspire for Rosie to be.

[00:09:43] Host 1: Paul Barnhurst: Yeah, yeah, I know, I think it's exciting. That's really, it'd be pretty amazing to watch. And it's coming like. Right. I know we all know this. The best chess players in the world have been beaten by machines in other areas.

[00:09:57] Guest: Dennis Jiang: Yeah.

[00:09:58] Guest: Dennis Jiang: Yeah. Yeah.

[00:10:00] Host 2: Glenn Hopper: You know, you mentioned the experts. I'm wondering for today what you're saying today. How do you see most people using Rosie AI now? And, you know, the most common use case. And I'm wondering because it's a problem throughout everything with generative AI. How do you manage the risk around hallucinations with a tool like this?

[00:10:21] Guest: Dennis Jiang: It's a great question and kind of a common problem across kind of like AI tools. I think the really great thing about how Rosie works, and kind of working with spreadsheets, is that you can always double check her work, you know, in the same way that when you work with an analyst, you know, like you ask them to like, hey, forecast Q3, you know, they build a model and they calculate things and they they send you an email saying, hey, here's the forecast, you know, but here's the model that kind of justifies kind of what I'm saying to you. The same thing with rosy. You know, she's doing all of her work inside your spreadsheet, inside of Excel. And so when she says that the forecast is 4.5%, you can trace back and, you know, check her work and collaborate with her and be like, oh, actually, that's not what I meant. Like, oh, I wanted us to calculate it this way. And she'll kind of like to change it for you. And, you know, that's really kind of like an illustration of how people are using Rosie kind of day to day. They're just using it in the same way that you collaborate with, you know, like a great analyst. And so, you know, people are kind.

[00:11:11] Guest: Dennis Jiang: Of.

[00:11:12] Guest: Dennis Jiang: Long, kind of like winding conversations with rosy, you know, like, you know, so, you know, start with a data set like, hey, like, Rosie, help me analyze this data, you know. And she kind of, like, works with you on the analysis. And it's like, hey, okay, like, great. We got to a good place, like writing an email to my boss who's, like, describing kind of what we learned here. You know, Rosie writes it, you know, and kind of like, and you go back and forth to that, and it's like, hey, like Rosie, I see, like this advanced formula that you wrote here. Like, explain to me how this works, you know. And so, so yeah. So we see people kind of using Rosie like literally all day at work. You look at their usage pattern, it's like literally eight hours every single workday. You know, just having this, like, just constant back and forth with Rosie, you know, conversation, you know, varies in the way that, you know, people use ChatGPT in a similar way. You know, it's like there's no there's nothing off limits. You know, like you're having conversations with ChatGPT and like with Rosie, it's a similar kind of thing, but much more focused on, like, the work at hand.

[00:12:01] Host 1: Paul Barnhurst: Yeah. Question. Totally off script. Again, I'm going to go off script one. You refer to her as a she'd a bunch of time when you started, was that deliberate or did it just over time kind of become a she? I'm just kind of curious. Right. Because some people will say it, some will give it pronouns and I just find it interesting how everybody kind of refers to it a little differently. So what was the kind of thinking there?

[00:12:26] Guest: Dennis Jiang: Yeah, a little bit of backstory here. So yeah, I mean, we've always thought of her as this agent. Agents are hot now, you know? But when we started a few years ago, you know, it was the future, but it was a little less hot. But the original name of Rosie was actually spelled with r o w c. So, like Richie Klein. So that didn't happen. And unfortunately, nobody could pronounce that when they saw it. It was like.

[00:12:48] Host 2: Glenn Hopper: Rousey.

[00:12:49] Guest: Dennis Jiang: You know. And so we ended up changing it, you know. But that was the original intention. So you know.

[00:12:55] Host 2: Glenn Hopper: Yeah. Funny side note I think I'm actually going to have where I own the domain for what will eventually be Rosie's first fan club. So I developed Rosie, the robo CFO, a couple of years ago, and when GPT rolled out, I was promoting other people's gigs and they were friends of Rosie. So I have the Earles, friends of Rosie and friends of Rosie I. So if you need a fan club website, you know, I think we, I think that I think we could transact around that.

[00:13:27] Guest: Dennis Jiang: Yeah. That's perfect. Yeah.

[00:13:30] Host 1: Paul Barnhurst: And there you go. When you develop all your additional friends. Got the website.

[00:13:35] Guest: Dennis Jiang: There.

[00:13:35] Host 2: Glenn Hopper: Yeah.

[00:13:37] Host 1: Paul Barnhurst: I'm curious, what's your thought as I continue to develop, at what level will people still need to learn Microsoft Excel? Kind of well, we get to a point where they don't need to learn it. I'd love to kind of get your thoughts there. As we start to see these tools, doing a lot of things we used to do in Excel is much quicker and easier.

[00:13:55] Guest: Dennis Jiang: Yeah, yeah. The way that I think about it is like, I think you'll always need to learn Excel in spreadsheets. And I think about it similarly to arithmetic, you know, it's like, when's the last time you did long division by hand? Probably when you were in, you know, five years old to ten years old. You know, but you still need to learn it in order to, you know, even though you can use a calculator. And I think, you know, spreadsheets are a similar kind of thing. It's, you know, just like, you know, in your career, like when you get promoted, you know, to, like, become a manager of analysts. And in finance, you spend less time in the spreadsheet, but you have to understand it and you're the CFO. You know, it's like, yeah, you're not going to spend nights kind of like building your models anymore. But the skill set is critical. And so I think you're always going to need to learn it. I think the nice thing here is, if you're going to be able to learn it much faster, with AI and you're going to be able to say be saved from, like, all the grunt work. You know, I think with, you know, my experience with spreadsheets and I imagine kind of like yours in the audience, there's so much value with the spreadsheet and the model that you build.

[00:14:53] Guest: Dennis Jiang: But like 90% of the work is grunt work, you know, is kind of the tedious stuff. You know, it's like, hey, like, oh, man, I gotta debug this formula. I gotta figure out how this works or fix the formatting, you know, like so. And 10% is that, that really valuable strategic piece. Imagine if you could just do the strategic piece and like someone else, Rosie took care of the 90% of like, hey, like, you know, like turn this like from monthly into quarterly or, you know, hey, like change, like the way that we calculate this kind of metric in this certain way. Like so freeing you the analyst and kind of like the manager of the analyst. Up from that kind of work. One way we think about Rosie actually is like, we're giving everybody, everybody a promotion. So now you're the analyst, like, you have a junior analyst who just has to be phenomenal at selling, who's just taking care of all that work for you, and you are managing this. This AI agent could do this work for you. So, yeah. So congratulations to everybody.

[00:15:45] Host 1: Paul Barnhurst: You know so if I use Rosie, do I get a promotion with that. Or do I get a raise with that promotion?

[00:15:50] Guest: Dennis Jiang: Oh, yeah. That's between you and your boss, you know? But, you have more responsibilities now.

[00:15:55] Host 1: Paul Barnhurst: Oh, that's what I'm used to doing. More responsibilities, but not more money. I was trying to fix both.

[00:16:02] Host 2: Glenn Hopper: But that's. I mean, it feels like that's where we're headed. It's like coding for Excel. And we have a friend of the family. Their daughter is studying computer science in college right now. And her mom was asking, is she just wasting her time? And I thought, no, I mean, it's great. The best vibe coders are the ones who can also write code so that you're looking at the output and then what's happening, and you can go in and fix it because you know where your loop is, your conditional loop is broken or whatever. So the skills to have them there, it's like, I think you nailed it with it's more like being a manager. And I really see this future of work where you come not just with your degrees and kind of the models you've built before when you're hired somewhere. But it's you who kind of come with your agents too. And if you've got these Excel agents that are helping you, then you, you are essentially managing these, bots. Yeah, or whatever to help do your job. And then it it's I think that a big part of the crisis is, well, we all had to learn. We all had to hack our way through coding or through learning Excel or whatever. What are we going to do with entry level people right now? And how are we going to expect them to come in as managers? But I think people are resilient. I think we're going to start seeing nobody. Nobody goes and gets a master's in finance or a master's in accounting because they love data entry and plugging in. Well, actually, I'm in the wrong crowd here, Paul. I was I say nobody loves plugging in Excel formulas, but I would I guess I would say present company excluded.

[00:17:33] Host 1: Paul Barnhurst: A normal person doesn't love it.

[00:17:37] Host 2: Glenn Hopper : Yeah. You are wearing the Excel jersey.

[00:17:40] Host 1: Paul Barnhurst: And I already am long ago, giving up the idea of being considered normal. So we're.

[00:17:44] Guest: Dennis Jiang: Good.

[00:17:45] Host 1: 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.

[00:18:51] Guest: Dennis Jiang: I know it's funny you mentioned vibe coding actually. Like one way that we think about using Rosie is actually what we call vibe modeling. You're just telling her what you want and she's just building it for you and you're just going back and forth. So the same way that people use cursors. You know, it's kind of famous. But, yeah. No, I mean, I think, you know, for, for younger folks, junior analysts, you know, like, they're using AI all the time in college, in their studies. So, they're, like, naturally suited to, like, be managing all of these A.I. to do all this work. And, you know, I think it is, it's going to be a natural kind of thing. And, and one way that I thought about it as well is like, you know, like I was talking to some of my colleagues who I started kind of like as a consultant as it's like, remember when they sold us the job and it's like, hey, you're going to work on these strategy and kind of working with like, you know, high level executives and all this stuff. And it's like, that's why they sell you the job. And then like, you go into the job and it's like 90%. Like Excel and PowerPoint, you know, and there's still that strategic part, you know, but it's like the minority of your time. Like imagine if the job was actually kind of the way they sold it to you. You know, that's kind of like the future that we see.

[00:19:50] Guest: Dennis Jiang: And I think it's an exciting future. It frees you up. And one kind of thing, actually, I was thinking about,was people don't realize this, actually. But before Excel spreadsheets were done by hand, you know, so, like, you really, it was like your job as, like a finance person to, like, literally like if you get asked like, hey, like what? Change. Run this scenario for me. It's like, okay, it'll take us two weeks. And like for people like going through and doing all the calculations, you know. And so actually there's a lot of angst when Excel came out, it's like, oh my gosh, all these jobs are going away. Like, you know, what are you going to do? I'm like, no, no, nobody thinks that anymore. It's like, oh man. Like it's like, thank goodness. Like, you know, I don't have to do this by hand. And there's more finance people, more accountants than ever. And, like, what are they going to do? They get to work on higher order problems. The strategic problem that you studied for. And, like, nobody wishes to go back to that, you know, so but, like, you know, we're all freed up. You know, and I think AI is going to be a similar kind of thing, kind of where you get to focus on the thing that is more interesting, kind of more impactful, and kind of like outsource a lot of like the tedious stuff.

[00:20:50] Host 1: Paul Barnhurst: Fun story. When you mentioned, spreadsheets right by hand, the old green ledgers. I interviewed a guy, Colin. He's in his 80s. He still does modeling and training of people, 1978. He's in South Africa, and he and a buddy were building a model, and they rented out the bank in downtown South Africa mainframe from, like, midnight to 6:00 in the morning and built a three statement model using punch cards. So it tells you how far we've come, right? I can try to build it with punch cards. And they went and presented it to the bank and they wouldn't believe them. The investment committee couldn't believe him that they actually built a three statement model using punch cards.

[00:21:29] Guest: Dennis Jiang: Yeah. Well.

[00:21:31] Guest: Dennis Jiang: What a feat of engineering. But yeah. Wow. That's it.

[00:21:35] Host 1: Paul Barnhurst: You know, just talk about that kind of journey. I have a short one minute clip where he tells the story. I interviewed him on a financial modelers corner and it's fascinating. If you get the chance, I'll send you the link, because I think you'd enjoy hearing his hearing, kind of the whole story. But, you know, as we as we talk about all this, I would love to know what are the areas, the most time consuming parts that you really see Rosie solving? Is it, you know, help with formatting? Is it auditing and debugging? Is it, you know, building out the steps of the model? I'm assuming there are some areas. So I'd love to kind of get your thoughts of, you know, what are the areas you are most seeing it used for? What do you think are the areas that save the most time today?

[00:22:20] Host 2: Glenn Hopper : Yeah, yeah.

[00:22:21] Guest: Dennis Jiang: I think so.

[00:22:22] Host 1: Paul Barnhurst: Core.

[00:22:22] Guest: Dennis Jiang: Kind of usage that we see, you know, is like this, like core, just like building models kind of iterated on models, you know, like rosy, like, you know, help me. Kind of like, you know, calculate this or kind of like, you know, figure out how to do this analysis, etc.. That's the core of what kind of people we use it for, you know. But the edge case is, you know, the outside of that is an enormous amount of, you know, just similar to trashy. You know, I think it's kind of a place that I see just like a ton of people getting value and just saving a ton of time. Is this actually like ramping up on models? you know, so it's like, hey, like, someone sent me this model and like, I'm like, I'm taking over this thing. It's like, man, it's a beast. Like I'm like, I'm going to, you know, normally have to spend days kind of understanding what this person was trying to do. The original author was trying to do, you know, really just come in and you just ask, okay, like, rain me up on, like, what's going on here? Explain. Kind of like how this, you know, this she works where the relationships are enormous, you know, amount of usage kind of along those lines. And then also, yeah, this kind of, kind of like auditing, debugging piece, you know, it's like, hey, like, oh, man. You know, in order to double check someone else's work or even my own work from a while ago, I was like, oh, man, I gotta like go through line by line, just be like, hey, like, you don't see any mistakes.

[00:23:27] Guest: Dennis Jiang: Like, you know, she can do it in a super human way. Um, kind of like that $200 million example I mentioned before, you know, it really is like, you know, having this Excel expert. You know the world's best. You know all the folks on stage in Vegas at the Excel world. Imagine that person sitting next to you and just like, no judgment, no, no anything. You know, just like whatever you need. Like I'm here to help you. And that's what Rosie is. And so that's how. And that's how folks are using her. And so, like, a lot of, you know, like core modeling kind of pieces, but also as a tutor as a, as a double checker, you know, as a friend, you know, like as a, as someone helping you write kind of like emails to your, to your, to your boss, etc.. The one thing that that's a funny kind of thing that that someone customer shared with me was, you know, they asked for us to write an email for them to, to describe the analysis they did together. And then they said like, oh, afterwards, like, oh, it sounds to really make it less sound like I. And so she wrote another draft that sounded a little more like added typo or something. That's how we see people use Rosie, you know, just like a broad range of, you know, the way that you use kind of like a great colleague as kind of like a second pair of eyes.

[00:24:26] Host 1: Paul Barnhurst: Love that funny, funny story to go with that is, I saw someone the other day they were sharing. You know, I often when you read it, you're like, okay, this sounds to me. There's now a website that's designed to take the content you got and try to make it sound more human. That's all it does. That's using an AI, which is very ironic, but specifically designed to look at patterns and try to match them to be more human. Like, I just kind of laugh. I'm like, so we have the AI, right? It. Then we stick it in another AI to make it sound like us. Yeah.

[00:24:59] Guest: Dennis Jiang: Sorry. Come full circle.

[00:25:00] Host 1: Paul Barnhurst: Yeah. Like where have we? What has become of us?

[00:25:04] Host 2: Glenn Hopper : But the only thing it does is remove em dashes. That's it.

[00:25:10] Host 1: Paul Barnhurst: oh. Yeah. We'll save that for another episode. Glenn. Yeah.

[00:25:16] Host 2: Glenn Hopper : As we've been talking in Paul and I've had this conversation a lot, and we've had it with other guests as well, because there's with generative AI, there's a million tools out there, and there's there's software out there, SaaS tools off the shelf, stuff that does a lot of what Excel does and tries to automate forecasting, and Paul and I keep going back to it feels even I'll do stuff in R or even Python, but my grounding is always Excel. And so many people are like that. And I'm wondering if you look at what Rosie's doing right now, does it make sense at some point? Do you take out the middleman and you're just talking to Rosie and you're not in a spreadsheet? I'm just. What are you picturing? Because you are. You're tied to the hip with Excel right now. But Excel, you're using it just as you would in another tool with Rosie. So I'm wondering, there could be an engine, a different engine, or weather. And I don't even mean Google Sheets, I mean just an entirely different deterministic platform behind it. So if you're with that, if you're looking ahead 5 to 10 years from now and all the we don't have to create the formulas anymore. And what financial analysts and, and other people in finance are accounting are going to be doing. And how many people are doing the job I don't know. Do you have sort of a vision of what all this looks like 5 or 10 years down the road?

[00:26:38] Guest: Dennis Jiang: Yeah, yeah, yeah. I mean, the way that we think about it is, you know, we think about Rossi as his agent. and so, you know, right now she's in Excel, and actually, soon she'll be in Google Sheets, too. You know, but, but, yeah, like, going forward, actually, like, we think of her as this, like, analyst, type of AI, you know, so, like, you could imagine just like sending Rosie an email and she'd be like, hey, like rerun the numbers or, you know, like, hey, Rosie. Like, pull the data from, you know, SAP and Salesforce and rerun our kind of like, financials for, you know, last month or, you know, like forecast next quarter. And so she's able to kind of like Excel and the spreadsheet are like the core, you know, as they are in kind of any finance team, you know, but her ability to, to, to work across your tools and also work independently,you know, in the same way that the CFO isn't like, you know, directing the analyst to, you know, hey, change this formula or whatever, you know, similar, like you can kind of ask Rosie to do the same thing. And so, I mean, that's the way that we think about evolution. And I think, you know, like that's where we're getting close to that, that point. But yeah, I mean, actually another point that you made that I think is really important is like, you know, I think a lot of tools have come, you know, a lot of you have complaints about Excel.

[00:27:45] Guest: Dennis Jiang: You know, it's like, oh, it's bad at this, bad at that it's so hard to use, etc.. And so a lot of folks who have tried to tackle this opportunity have been like, hey, like, let's build a thing that replaces Excel. And I think that's a fundamental conceit. You know, that it's just like a nonstarter.And so kind of with Rosie kind of we start with places like, you know, as much as people complain about Excel, like, people love Excel and rely on Excel and kind of use it universally in their organization. So even if someone wants to change, like, you know, your organization doesn't want to reinvent a 40 year process and like, you know, redo everything. And so kind of the starting point for Rosie is like it has to work within the spreadsheet, you know, like and like whatever spreadsheet someone's opened, you know, like Rosie needs to work there. We don't need to ask someone to change. Kind of like, you know, reformat your thing in this way or kind of follow this format or kind of do things the way that we want you to, like, we want to adapt fully to the user. And that's part of the magic that makes Rosie so compelling for our users.

[00:28:34] Guest: Dennis Jiang: Like, hey, I didn't like it's just like, so natural. Like, I just, you know, whatever I do. And like you, we've discovered, like, so many ways. Like, I mean, I knew that people used Excel in different ways. I was just like. And the universe is way, way larger. You know, but we are, you know, we're rolling with that. And, you know, our lives would be way easier if we told people like, hey, you have to do it. Use the spreadsheet a certain way, you know? But like, we know that users don't want that. Some organizations don't want that. And so kind of like a real starting principle is like the spreadsheet is not going away and is like really the foundational piece. You know, we build on top of that. And we think of spreadsheets. You know, even if people aren't learning, you know, like how to do formulas, kind of like day to day anymore, like the spreadsheet itself is going to be an important form factor going forward. You know, it's the way that people collaborate on business logic, really, if you think about it fundamentally, and it's just like a way, you know, the way that you collaborate with the AI is like through the kind of like expressed in the spreadsheet. So we think spreadsheets are not going anywhere. And we built kind of like that with that in mind.

[00:29:31] Host 2: Glenn Hopper : Yeah. Great answer. And that really, it's funny when you think about just how those of us who've been, whether building them or reviewing them or whatever, those of us who've been in spreadsheets forever, your brain just starts to work and excel where you've got your rows and columns and your different tabs, and it all makes sense and everything. It's orderly and organized versus looking at a database schema and going and trying to figure out all that. Excel just keeps it up in the spreadsheet right there in front of you. And so you can understand the data and trace the links and all that. And it's a good visual representation of it. And I think business minds, certainly for anyone working at this time, you came up with Excel and it'd be hard it'd be hard to sever that.

[00:30:14] Host 1: Paul Barnhurst: It's a very simple form factor. At the end of the day, it's easy to understand, easy for prototyping, you know, regardless of whether it's Excel, Google Sheets or one of the other 20 new spreadsheets that have come up. Right. They're all still using that basic idea. And then you have those spreadsheets that stick a database on the end and go to multidimensional modeling. They're still a spreadsheet in many ways, right? Even though almost every budgeting modeling tool that's tried taking people out of Excel, very few have said we're not going to use the spreadsheet. At the core, you have a few that have done visual, you have some that have, hey, we're going to do inputs and then display it in different ways. But, you know, 90% of them are still really some kind of spreadsheet form factor.

[00:30:55] Guest: Dennis Jiang: Yeah, absolutely. Yeah. There's a joke I've heard in the BI business intelligence space, where the most popular feature of every product is the export to Excel button. You know, because it's the universal kind of form factor that every tool is stitched together by exporting to Excel and importing to the next thing and then exporting back into Excel. And we think that that's unsolved. Like there's always the human with Excel, like sits in the middle of all these things you've got, you know, some kind of wizard who is just like keeping the business running, you know, by being kind of like this universal layer and kind of like we, see Rosie kind of playing that actual role and kind of like being able to, you know, to bring kind of like a lot of, power to kind of like, you know, with the spreadsheet as the, as the centerpiece of that.

[00:31:37] Host 2: Glenn Hopper : That's so funny. You said that every BI dashboard I've ever built, the first question I get is, where do I export this to excel. Yeah. Yeah. Exactly.

[00:31:46] Host 1: Paul Barnhurst: Yeah. Well, the way I say it, the three most popular buttons in programming. Okay. Or enter, you know, escape, slash cancel and export to accept. Yep. And I'm not even sure it's in that order.

[00:32:01] Guest: Dennis Jiang: Yeah.

[00:32:02] Host 1: Paul Barnhurst: Yeah. So there we go. All right. We're going to move on to our fun little section we have here as we get closer. Wrapping up what we did is we took your bio, your online information, the questions we developed today. And we asked ChatGPT in this case to come up with 25 fun, unique personal questions we could ask you. Okay, so here's how it works. I take one approach.Glenn takes another because Glennis a little more eye focused than I am. So you get two options here. Option one you could pick a number between 1 and 25, and I will ask you that question. Or the random number generator can pick a number between 1 and 25. And I will ask you that question.

[00:32:44] Guest: Dennis Jiang: Let's go with the random number generator.

[00:32:47] Host 1: Paul Barnhurst: All right. Here we go. I'm going to hit it between 1 and 25. I came up with one I don't think I've had before. All right. So this is under the section of Startup and Vision. And I think you can answer this a little bit, but we'll go ahead and ask it anyway. What was the exact aha moment when you realized Excel needed an AI assistant like Rosie?

[00:33:13] Guest: Dennis Jiang: A great question. For me, actually, it's interesting. It was the moment that I got access to GPT four, which I think was in March of 2023. And so I've been playing around with LLMS for a while. You know, like when we came out and before that. But it was when GPD four came out, I got access to it. I was like, hey, this is really good at writing code. You know, which is great. And I'm like, if it's good for writing code, it's probably good at writing. Working with spreadsheets, actually, because, you know, I mean, honestly, like coding and spreading, like working with spreadsheets are similar kinds of thought patterns, actually. And so I discovered, oh, man, this thing's really good at spreadsheets. And it was a lightbulb moment for me. I was like, wow, I know what's possible now. That was not possible before. The thing that so many people I've worked with and myself like, have needed, like, our whole careers. It's just like, you know, an AI that can help us, kind of like, navigate the challenges of using, using spreadsheets. And so, yeah, so, I mean, really, that was the kind of starting gun and kind of Rosie has been on this kind of journey ever since that at that moment. But it was very clear. Yeah. Sometimes it's a little hazy, kind of like when, when the inspiration came, it was very clear for me, actually.

[00:34:23] Host 1: Paul Barnhurst: 4.0. Got it. All right, Glenn, you're up.

[00:34:26] Host 2: Glenn Hopper : All right, so my approach. Since I created these, I just copy and paste them. If Paul holds the questions, I just copy and paste them in and tell him to pick one for us. And Paul said GPT did a better job with the questions this week. It's that some of the I don't know, maybe, maybe.

[00:34:42] Host 1: Paul Barnhurst: No, I don't use the prompt every time I make it up as I go. So that could have a little bit to do with it.

[00:34:47] Host 2: Glenn Hopper : We need your prompting. All right. Let's see what we come up with from ChatGPT here. Oh, we're. You know what? I got number 25. You got one? I got the Alpha and Omega here.

[00:34:59] Host 1: Paul Barnhurst: I like 25. This is a fun one.

[00:35:02] Host 2: Glenn Hopper : Statistically, how likely is it that we're going to get to, I guess no more likely than any other two numbers. That's a fallacy, I guess. But anyway, we're.

[00:35:11] Host 1: Paul Barnhurst: Now that we've had our, false statistical course, we'll move back to podcasts.

[00:35:16] Host 2: Glenn Hopper : Yeah. All right, I like this. So, let's see if AI was featured in a sci-fi movie. Would it be a hero, a sidekick or something else? Don't say villain. I don't think the villain is coming to take your job. I don't.

[00:35:33] Guest: Dennis Jiang: Yeah.  Well, I'd like to think of her as a very important sidekick, perhaps, you know, but, helping the protagonist navigate some difficult analytical challenge. And, you know, pairing with them to, to get to the, the brilliant insight that kind of solves the case and kind of, you know, climax of the movie.

[00:35:54] Host 2: Glenn Hopper : So more like Robin to Batman than Aquaman to Superman. I don't know Aquaman. I don't know what. Yeah.

[00:36:02] Guest: Dennis Jiang: Yeah, yeah, I, I'm less hip on the superhero kind of like law. But yeah, some, some super smart, you know, like, maybe James Bond is like, you know, like, what's the person's like? Their, their, their side. That was that. Yeah. Yeah. Their genius sidekick.

[00:36:21] Host 1: Paul Barnhurst: Love it. Well, you know, thank you. Thank you for joining us, Dennis. As we wrap up, I know you mentioned and we'll make it available in the show notes and offer where people can get, you know, one month free access to Rosie if they want to test that out. So we'll put that in the show notes for people and appreciate you carving out some time and chatting with us today. And good luck with Rosie Eye as you continue to build and grow. I know it's an exciting time and a lot of work when you start a company.

[00:36:47] Guest: Dennis Jiang: Yeah, yeah, thank you very much. It's been such a pleasure speaking with you.

[00:36:51] Host 2: Glenn Hopper : Thanks, Dennis.

[00:36:52] Guest: Dennis Jiang: Yeah. Awesome. Take care.

[00:36:54] Host 1: 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.