There may be errors in spelling, grammar, and accuracy in this machine-generated transcript.

Isaac Heller: Hey, everyone, this is Isaac. I'm CEO at Trillian. And today on AI, accounting, intelligence. We'll be talking with real people about real things related to AI and the impact on the accounting industry. Stay tuned. Good day everyone. I'm here with Nick Boucher. I'm really excited to talk to Nick again. We've got to know each other, you know, in recent months. [00:00:30] Um, Nick, good to see you again.

Nic Boucher: Yeah. Thank you for inviting me and talk to everybody who is interesting in interested in AI and accounting and finance in general.

Isaac Heller: Very nice. Well, Nick, I'm from, uh, Texas, and I did hear the name bullshit when I was young, but it was from my friends in Louisiana, from the Cajun, the families in the South. Bullshit. Where are you? Where are you from? Where are you? Where are you talking from today?

Nic Boucher: So I'm living in Germany, but as your [00:01:00] audience might hear, I'm coming from France. I also lived in the US one year. That's where I met my wife, who is German, and now together we live in Germany. I worked also in Luxembourg and Singapore when I was working at PRC. And basically, yeah, I feel myself as a citizen of the world and a happy also to be able to talk through technology to everybody in the world, because I think we are really lucky right now that we can have access to people like you, like [00:01:30] in the US or in Israel. Um, I think later I will talk to people that are in the east of the world. So happy to, uh, to speak to this worldwide audience.

Isaac Heller: That's amazing. And I think that's why you and I get along. You know, I, I grew up in Texas. Of course, I spent a little bit of time abroad in southern China. My wife is also German, which is which is fun. And, you know, we've lived in a few places as well. So it's really exciting. And I know today we're going to go deep [00:02:00] and wide in terms of international versus us and startups versus enterprises as it relates to AI, innovation and finance. But before we jump in, I mean, like, Nick, you've built a pretty cool career out of this. You mentioned PwC, but maybe just tell the the viewers and the listeners a little bit more about how you became really one of the experts in innovation in finance.

Nic Boucher: Yeah. So it started actually, uh, as a kid, I always love to use technology, uh, for [00:02:30] to to help myself to, um, also like, as an anecdote, my dad when I got my first computer. So at the time almost nobody had computers and I had a computer. And when I was on holiday instead of, you know, like playing games or, uh, playing, um, outside, he asked me, first you need to do for me some work. And so, you know, like all of these Excel templates that you need for checklists and all of this. My dad was drawing that on [00:03:00] paper and asking me to do that. And he was really demanding because, uh, he's a technical draw. So he really wanted that. Everything is perfect and aligned. So that's how I learned how to use Excel, how to use the computer. And later on I use it also for fun. But really for me, it started when I was at PUC and I was looking at everybody working really hard, and I had also to work hard. But I found that if you had the right Excel formula, if you knew how to use Excel [00:03:30] the right way, and after PowerPoint, I could just like leave and be done before everybody else and or do more than the rest. So and then what I loved about that is not keeping that for myself, but also sharing that with my teams. And later on, I felt that there was much more to share than only for a small team of ten people. So that's where I went on and go on LinkedIn to have like a bigger, um, bigger speaker. And now when, [00:04:00] uh, AI and ChatGPT really got democratized and more than one and a half year ago, I really felt that both what I've built as an audience, plus my knowledge in finance, because I worked 15 years in leadership position, plus my knowledge of technology and my curiosity, I could combine all of the three to show to a lot of finance professionals how to use AI in finance, and especially generative AI.

Isaac Heller: That's awesome. And I must say, we have to put you in the category [00:04:30] of influencer. Given a lot of the great stuff that you've put out and people read. And I really love to hear that you talk about kind of the curiosities you had when you were younger and with your father and stuff. And I think that, you know, there's a lot happening in the finance and accounting industry, and sometimes people feel a lot of career pressure. I've always found that sometimes it's good to think back to what you enjoyed when you were, you know, early teenager or maybe even when you're a kid, uh, and just kind of continue in a place that will make [00:05:00] you passionate like that. What I've also noticed is that there's a few people like Nick who are putting out great materials online, whether it's, you know, templates or resources. I mean, is there like a community, would you call it finance and accounting influencers? Is there a community? Do you have a lot of friends in different regions or different segments, and you guys have gotten to know each other over the years?

Nic Boucher: Yeah, we, uh, that's how myself when I wanted to work and build my own business, I reach [00:05:30] out to people who were already, uh, giving training or having their own consulting company or doing fractional CFO. And I kind of build my circle of influence, uh, with people like Anders Lindbergh, Josh Aronoff, Paul Barnhurst, and because they were like one step ahead of me and, you know, like you are the average of the five people you talk the most to. And so for for a while, we were really like, we had a really small group where we exchanged a lot of best practices. [00:06:00] And now as there are more and more people, especially on LinkedIn, sharing information, but also people giving courses or people doing consulting and finance, I build a big slack group with everybody that wants to join from finance, where we really share a lot of best practices because all of us, we are either like working alone or having a team of freelancers with us, but we don't really have colleagues. And like this. We are a group of 50 people with colleagues and [00:06:30] it's crazy to see, like the value shared inside of a lot of people. I think they they gain like six months of time just by joining and asking questions and being ahead of everybody else who is not in this group.

Isaac Heller: That's amazing. And it's connecting the world. And, uh, Nick mentioned a few names like Enders and Josh and Paul and, well, those are all great resources as well. I think Paul, uh, at Barnhurst, right. That's his name. Yeah. He's known as the Fpna [00:07:00] guy, is that right? Something like that. Okay, so Nick Boucher, fill in the blank is the blank guy. What is what is the the area that you're going deeper into in this big world of finance and accounting?

Nic Boucher: Yeah. Well, I would say for me is Nicola is the I. Finance guy because that's where I am spending my time today to research a lot, to connecting, to connect with people like you, [00:07:30] with founders in the, um finance, tech, AI industry. And I also built the AI Finance Club, which is a community and membership of CFOs, fractional CFOs, head of finance who have, you know, like an appetite and a curiosity on how to use AI in finance. But they found out that by themselves they cannot learn everything. And also they have a job on the side. So even though it's fun to learn, they [00:08:00] there is just too much. And by putting everything I know and everything I'm working on, but also also just for myself, it's not, uh, enough. So I'm bringing experts like Christian Martinez, like Glenn Hooper, like, um, Emma Schipper, the Excel dictionary girl who is really good with AI and Excel. She's also coming as expert. And I'm gathering all of these experts who already spent a lot of time working on how to use AI for finance [00:08:30] and putting that in an exclusive environment where every week people get either a video course, an article, or a masterclass on how to use AI for finance.

Isaac Heller: That's awesome. Um, and just for those of you like, uh, viewing or listening, it's the early days of finance and innovation, you know, I mean, if I would have maybe gone back, I don't know, 30 years ago or whenever Excel was coming out, I guarantee [00:09:00] you there were just a few people using Excel and early Excel, and there was clubs and groups and, you know, now it's become ubiquitous across the world. And so some of these skills, which I know we'll talk about later, uh, are still in the early days, you mentioned, um, you mentioned founders. I mean, first of all, you know, me again, Isaac from Trulia. And like, I, um, we love to talk about AI. We love to talk about finance. We love to talk about accounting, but we [00:09:30] tend to talk about the things that our software does, like leases, rev rec, audit, whatever it is, I, I think it's really important that the leaders of the next kind of movement of AI and finance are practitioners. People are coming from within. People have experienced themselves. Obviously, you do partnerships with, I'm sure, good software and AI and stuff like that.

Isaac Heller: But like, I think it's so important that the next generation is practitioners. You know, the software providers will be there, but, you know, you [00:10:00] need to hear from the people that can actually show you day to day, not just what they're doing in specific areas, but kind of across the spectrum. And you, you know, talking about founders, I remember your eye map when it came out about a month ago. So if you guys haven't seen this, uh, Nikolai Nicholas has a top 100 AI finance tools. This had over 1000, you know, tons, tens of thousands of views and reactions. I also realized that just in the past couple of weeks, the [00:10:30] biggest in one of the most prestigious venture capital groups in the world, Andreessen Horowitz, came up with their own AI and accounting map. I mean, first of all, what's your reaction? Because it's not just Andreessen. I mean, Greylock wrote an article and Emergence Capital wrote an article, and we're seeing more and more reaction there. What's your just your initial reaction when you see all these big guys a little bit after you creating their maps?

Nic Boucher: Well, I'm if you have to compare, [00:11:00] like my structure, where I'm alone with a team of freelancers and some experts and Andreessen Horowitz, I'm quite glad that we could release that. And ours is actually more complete because I think they have maybe like 50 tools. We have 100. Uh, I don't really compare myself, um, or will not compare both, because what I expect from them is that they have had, uh, really deep analysis on each of the tool when [00:11:30] I was building that. Like, there are maybe like a list of 300 tools behind. And I was really, like filtering all of the tools to see, okay, which ones already have clients, which ones have serious funders, which ones have use cases that that are really I and not just like a mini mini layer, which was really hard. And um, for me, there was a lot of credibility about my voice when I was building this AI roadmap. Uh, this AI map. Sorry. And I think we [00:12:00] spent like, like from the beginning, I think it was like six months ago. I did a list when I started the AI Finance club and then really like at the end to select the 100. We we took another month then to build that on. Also visually it takes also a lot of time to group that. Then the grouping was really hard and I remember like 4 trillion. I was not sure if I should put that in revenue recognition and lease accounting in accounting automation. So and I ask you after and then like there is like still some interpretation [00:12:30] to to correct.

Nic Boucher: But this is really hard to do because I was my first AI or my first tool map, and I think Andreessen Horowitz, they have a methodology to do that. So I had to learn how to do it. But for me, what was really important is to talk to the people behind these tools. And I think out of the 100, I think like half of them, I had a call with them and I did the demos and all of this to really vet what was on the on the map. [00:13:00] And I think if I did that, it's not really for me. I did it because I was asked almost every other day, oh, which tools exist there and which tools are serious? And then I thought, okay, I need to do that because, um, there is a demand, there is a need, and And I see myself as somebody who has a chance to talk to all of these founders and to see these tools, and I wanted to do that for all of the people in finance, and also to give some exposure to the people that [00:13:30] are doing a great job like yourself without like, this is there is no real partnership behind. But I think if somebody is doing a good work, I need to use my audience to show that the good work is done. So like this. The word in finance gets better.

Isaac Heller: That's awesome. And yeah, I mean I think that's, that's a, that's a really important emphasis is there was demand for it. I mean when people are trying to build their voice, uh, even their career, and they want to go on LinkedIn or they want to go on Twitter, [00:14:00] they want to go on these sites and build it. It's great to be a leader, but at some point you have to understand what people want and kind of construct to some extent your content, your message, your education around what people want. And it's just the same as it's always been. You know, if people are asking 100 questions and you could give it to them in one place, that's successful. Now, Nick, I'm just thinking about all those startups and what number 101 and 102 and 103. I remember when I was starting four years ago to [00:14:30] to be on a map or to not be in a map. It's like, oh no, why didn't you why didn't you pick me? So I'm sure there was some, some tough ones that didn't make the map or were very close. Do you plan on updating it? What's your methodology for keeping up with the top 100 tools? Does it go to 200? Does it get refreshed every year? How are you thinking about it?

Nic Boucher: Well, the big advantage of having done this, it's, you know, it's the first version. So it got me into a lot of discussions with either tools [00:15:00] that were not on the map and asking, oh, um, which we should talk because I think you will like our tool and we need to show it to you. So like this. I don't have to search for it anymore. Like tools come to me. So that's a big advantage. The second one is also that I talk to people in the VC environment that I have already, um, skin in the game and, you know, like they talk to a lot of early startups. And recently I talked to somebody from the [00:15:30] PayPal venture capital fund, and that was really interesting to to hear the other stories, to hear maybe more about fintech than only purely accounting. And I And that's for me, an opportunity to learn more tools. And each time I learn more about the tools and the industry, then I can give that to the audience. And knowing the fact that more than 1 million people follow me, that's 1 million people more that will get to know about these tools.

Isaac Heller: That's awesome. And I do think [00:16:00] the perspectives are important. It's really hard to say. I'm going to do my best and put the 100 together. It's very I would imagine it's a very vulnerable feeling. Right? Because these are all great companies and they do something different, and maybe the perception of their company is in the wrong, you know, and to put that out there, it's really important. But what's more important is doing it, you know, I mean, we need Nick's perspective. We need andreessen's perspective. We need all the different perspectives and kind of to put it together. Um, maybe I'll ask you a bit of a curveball. [00:16:30] I'm looking at your your finance chart, and, you know, a lot of these tools have have raised funding in 2024. I know that numeric just announced a round, which is a closed management tool. Clarity is a document reading tool. Just announced a big round. I know that Data sniper over in Amsterdam announced a big round as well. I think they made it to unicorn status. I mean, I could go down. There's so many, um, companies that are announcing, [00:17:00] uh, new rounds. Gebbia is another one that's more like fund research and, you know, just announce 130 million from, you know, Andreessen Horowitz. Do you have a perspective on that early scale venture model? Do you think some of these tools that are raising a bunch of money are complete game changers? Do you have a sense of is this overhyped under hype just right? What's your what's your perspective as you see the dollars coming in with the maps? [00:17:30]

Nic Boucher: So what was interesting is I got to I get to see like really young tools, I would say tools that are just developed where you see that the function is not something new, it's just either the UX is better or the niche down on one topic. And like one example, a lot of tools now are just tools where you can query the data. So they ask the user to load structured data. [00:18:00] And then there is like a chat bot where you can talk with the data. And for this, like having now seen a lot of tools when somebody showed that to me, I am a bit suspicious if it's a real innovation. Uh, so I take a bit like I am patient about that and I want to see, okay, like who is really using it? Like, how are you going to deal with the problem? Number one is that if you already have a structured data, probably you have already a business intelligence. And then the value added by querying [00:18:30] the data is not really high. It's more if you come and do something from unstructured data, then you can bring a lot of value. And that I almost didn't see, like the black box, you know, like the magic black box that will unstructured data make them structure and then create a management, reporting or insights for the CEO that almost doesn't exist yet. Or if it exists, it exists maybe just for one vertical. And so this tool I'm I'm kind of still [00:19:00] waiting for that, is there? What I see is tools like yours or like puzzle or. Um, yeah. Like numeric um, or like linear or or absent where they are really good in one vertical, meaning that, uh, if you give like a workflow or, uh, yeah, like the value chain from ah to, from A to Z, then everything is covered.

Nic Boucher: But what [00:19:30] actually like a lot of small companies want is the magic tool, magic AI tool. And they want that because they don't understand yet what AI can do. And they got promised that AI can solve everything. And once you go beyond that, like once you educate the the CFOs, the CEOs, they understand that they have to focus actually really to vertical and to take what is existing now or to take also open source or or off the shelf products. And I would say like [00:20:00] big tools that are going to do much more than what is currently available is yet to be done and is much more work. And maybe that's my prediction, but once Gemini, um, OpenAI, uh, Microsoft Azure with OpenAI and anthropic, once they are going to have APIs with all of the existing tools, that's where it's going to be hard to compete with this, because even though they are not specialized [00:20:30] in finance, they just need to have an API with QuickBooks and an API with the CRM tool like Salesforce. And then they can they can create a lot of value or make a lot of automation. And that's, I think, really hard and dangerous when you want to build a startup right now, because this is either you benefit from it if you know how to use it, or either like it kills your business.

Isaac Heller: Got it. So that by the way, that was great articulation. I'll share, you [00:21:00] know, kind of some of our experience as well. So when it relates to the funding. So a startup announcing 10 or 30 or $50 million out of the gate from a big investor like Andreessen Horowitz, um, presumably, I mean, we are finance guys, you and me. Presumably they believe that that can be $1 billion outcome or even more. And realistically, in order to be $1 billion outcome, you have to do something amazing or compete in really big spaces like ERP, Fpna [00:21:30] and such, or have just multiple products and grow. And I think what you said, you said like magic wand or magic box, like I think that that's where they're expecting those companies to be one day when they invest them in. So they're thinking today it's it's one check. But in ten years or five years or whatever the horizon is, it'll be a magic, magic wand. Now, Nick, we see what you see, which is there's not going to be this magic wand, right? We got all of the AI information. [00:22:00] It was very exciting. There was some very new innovative use cases, but a lot of the challenges that we have in finance and accounting are the same today as they were yesterday, which is getting a complete data picture. And I think that's where your point about the the structured data, they're called like wrappers, you know, like a, like a GPT wrapper on top of structured data.

Isaac Heller: That might be cool. It might be impressive. We call it a party trick. It's like, hey, look what I can do. I downloaded ChatGPT and put put some finance data in it. Um, [00:22:30] but the challenge is not necessarily that you call it business intelligence or insights. The challenge sometimes is getting all that structured data in the right place, so that the LLM or the GPT can, can bring it in. So I think you're definitely right that those wrappers are going to look really cool right now, but they're going to fade away or be tucked in to a broader feature set. So for example, um, I saw a startup that that automates technical memo writing write memo writing. Well that's a that's content creation. [00:23:00] That's a really good GPT use case, right. As long as you have the data. And so I would think that memo writing shouldn't be a standalone tool. It should be embedded and maybe ERP, fpna, whatever it is, some of these things where if you have the data, you can create it. So that's kind of the the rapper idea. And then yeah, you you got it on the workflow. I mean we, we at Trilion, we took the perspective of number one, in order to do the data extraction, you have to have something unique. It has [00:23:30] to maybe you know, what we can do with leases or revenue contracts is proprietary.

Isaac Heller: Now, LMS may be able to do 70% of it and keep going 1% every year. But we're going to stay five, ten, 15% ahead because we focus on it and our clients value that incremental value. But that's even that's not enough. Even that's not enough, right? Because Gpts can be cheaper than buying a new SaaS license. And so then we have the workflow. So it's like GAAP or AI for us. And then you connect the data extraction [00:24:00] or the workflow. Then maybe you have something, but just talking that out loud and talking with you, I realize like you really have to compete hard to get ahead of the GPT rappers and to stay ahead of Gemini and OpenAI, which could be an API into into the tools. So I think I think you have a really, really, you know, good perspective of the market. You mentioned, you mentioned the the Gemini's and the OpenAI's of the world. Do you think that they're [00:24:30] going to catch up and take more market share of the eye finance market, or they are going to be these kind of emerging new tools in the AI space? Do you have a sense of or your crystal ball and whether the big guys take it or some startups have some room to grow and you know, you know how we feel it truly in awe our customers. Yeah.

Nic Boucher: I think for big structure, I believe they will still count and they will need ERPs. So [00:25:00] the question is, how fast are the ERPs going to have really AI functions to compete uh, with like if you see what you can do is now a GPT, you are expecting that from SAP from Oracle. Um, they are starting like they are cooperating. Um, for example SAP is cooperating with Microsoft Copilot to to, uh, to work with in office and maybe tomorrow. That's what I expect, is that by the end of the year, you can do a bank reconciliation [00:25:30] and send an email to your clients with a connection between SAP and Outlook and Excel. Uh, but they are already not yet like they promised this, but we are still yet to see if it works What I what I think is for smaller structure. If today you can use something with Zapier, with even Power automate from Microsoft just and document intelligence for Microsoft just to [00:26:00] get the invoice, then extract the information from the invoice and book this inside your QuickBooks or Xero or NetSuite. Then why do you need a OCR tool for your accounts payables? And I think this is going to be for a lot of structure, like the the arbitrage between, okay, we are as a company level going to automate with these three four tools.

Nic Boucher: Or each department will take care of their [00:26:30] own, um process and choose the tools they want. And I think as it would be a choice between like how strong is the department against, uh, the company's central automation team or process team? But I think if the, the ones that are going to be faster at developing those automation are going to be OpenAI, Azure, Gemini, and maybe like the [00:27:00] ones of Zapier, the ones of, um, the automation tools. And for this, I think the IT team is going to come and say, well, you don't need, uh, an account payables tool. We can do your, um, your account payables right now with just like it was within one day. And it's really true. Like you connect document intelligence to your QuickBooks and then it's done. It's like a really easy API integration. And you can take a lot of processes like this and integrate between departments. [00:27:30] And I think then the current tools are risk because the more the easier it is. It is to connect the generic tools like Zapier, OpenAI, Azure to the ERP tools or CRM tools. The hardware is going to be for finance tools to exist, especially at the cost where it is right now, where it is big SaaS cost, and you will have to have an edge to stay and to be [00:28:00] relevant.

Isaac Heller: Yeah, definitely. And the even circle back, you know, we live in the venture world and we've, you know, truly and we've we've started to build a more efficient business in terms of our operating model. But a lot of my peers, when they take the capital, they operate at a very, very kind of high burn rate. And I think that as cloud costs go up and AI is expensive, like processing or processing costs [00:28:30] and tokens and all these areas are very, very expensive. And so if you're just passing through document processing, those margins are not the same as classic SaaS companies, right? You're your Cogs. I don't think people are accounting for it correctly in terms of a, you know, kind of AI costs in their Cogs. I'm sure they they'll get there. But like, it'll be really, really hard to compete with incumbents who are already installed. And you'll want to raise your SaaS [00:29:00] fee. But people will say, okay, well, I don't need to pay that much for an additional tool, right? So I think it's very, very interesting. Now you've you've talked about different sized companies having different roadmaps. Right. And you can imagine someone who's listening, maybe working in a startup, maybe working at a privately owned family business, maybe working at a public, uh, or large enterprise companies. How do you think about, you know, there are a couple different categories. [00:29:30] What are the different types of roadmaps that different sized companies are thinking about?

Nic Boucher: Yeah. So I for me, there are three types of roadmaps based on the size of the companies and the smaller companies could be also the same type of roadmap that you could have in a team of a big company. So let's start with, uh, SMEs. So small size companies there. I think you have the flexibility, [00:30:00] but also the constraints of low budget and also the constraints that you have. You don't have that many tasks that are high repetition with low value added, like everybody works in a lot of different tasks. So if you want to to save time, you have to work on a lot of different tasks. And for this, I will first advise for the team to be upskilled on how to use generative AI and also how to use machine learning [00:30:30] if you want to do forecasting with Python. So those are the two skills really to build this know how to use ChatGPT, Gemini, and even copilot in your or all of your tools that you already have today, and see how you can do your task better, faster so you can produce more and work less, or at least produce, um, more with the same amount of time Then, uh, the second part is, if you are already, [00:31:00] let's say, a medium sized company, and there is one process that is really demanding. So let's say accounts payable, um, or let's even say like for your if we take your industry or your niche, where is this accounting? You are a company with a lot of contracts for leasing. Uh, maybe like either you rent out, um, uh, like machines or you borrow machines.

Nic Boucher: But if you are in this industry, then you might need for one of the processes [00:31:30] or for example, for lease accounting, a specific tool, uh, like a trillion. So for this, if you are medium size or if you are in a big international company, but you have a business that is focusing on that, then I will our advice to start getting specific AI tools that will accelerate those processes and allow more time for the people working on these processes that are really demanding in time, but with low value. And also because [00:32:00] those processes will bring the data more structured, then normally the management and the people working on it will have faster and better insights. Then what you can also start is to start also using models for forecasting, for machine learning that are existing because you have enough data, because if you are like a medium size, you can start to combine macro and um, and internal data to do some forecasting on existing models [00:32:30] like time series or linear regressions. And this is not so hard anymore to do, because it's not only for data science. You don't need a data science team. If today you use ChatGPT and Python together, you can really like in half an hour, build a mini model that shows you if you increase your marketing expenses in some areas. What is the impact on your revenue? So that's I would say like for medium size and then for big [00:33:00] companies. So inside their companies, the teams of these big companies, they can do like the SMEs or everybody will upskill their themself and use the open tools or the off the shelf tools.

Nic Boucher: And inside those big companies, either countries or I will say business units can do like middle sized companies, but at international level and at really like, uh, headquarters level, you can start having teams that are really focusing [00:33:30] on using AI, uh, within all of the data that you have, like having a big data lake where you connect your CRM data, your ERP data, your supply chain, um, your web analytics data, and where from this you are going to also build your own models to get insights, but also to forecast. And like I think nobody in this company is weighted that OpenAI was there or uh, Gemini was there. They are. They already have teams working [00:34:00] on that. But now those teams first they get either more means because everybody knows that it's a priority or they get more pressure because everybody is also expecting more from them compared to before. So I think that's one. And second, you will now that now that llms are cheaper and also easier to use, you will have to start to make value from all of the data that exists inside your company to have an edge compared to the data that [00:34:30] is outside and available for everybody. And that's where again, like you will have to combine your own data with the, the, the LM models that exist to make your own model where only you have access to that and you can create value from this.

Isaac Heller: Um, I love that. And, you know, if you just heard that, you should probably rewind and listen to it again, because those were very good, clear recommendations for kind of the three different levels of, of businesses. [00:35:00] Um, you mentioned one thing, Python, which I want to come back to. Right. That's a scary word in some circles. Right. That's for engineers, not for finance professionals. But maybe you'll you'll correct me. Um, but also, it sounded like it sounded like when you can implement some of these, uh, AI models, whether it's at an SME or a large company. You talked a lot about insights and forecasting, right? A lot of what we do at Trillian is accounting, right, [00:35:30] and accounting compliance. But we always say, like, if you're getting the data in for GAAP and IFRS to streamline it, you may be able to do more with it. I guess my my first thought is, do a lot of the AI initiatives get driven by cost savings or more insights, or is there some mix? Um, first of all, do you have a sense there of whether people are driven by cost or insights or both or something else?

Nic Boucher: Yeah, I think that generally is really hard to [00:36:00] identify the cost saving, because it will not like make a team obsolete or a team, uh, irrelevant. It will just help one part of the process to be either faster or to produce more value. And people are still there, and you just need to use them better to also from themselves, create more value. And that's why because in finance we are not really like if you take [00:36:30] accounts payable apart where there it exists already since a long time that you have automation of um of invoice processing I think there. Yes, you can save cost, but for all of the other processes, I think there is a lot more of creating insights, getting better forecasting models, and also getting people out of the data crunching and that they spend more time on analyzing the figures, but also just, you know, use their human [00:37:00] skills to go and talk to the business, to go and figure out themselves, uh, some propositions to help the business. And also maybe for some teams, just have more time for themselves as, uh, as individual. Because having overworked team also is not good for the business. So if you can have all of this together, it's really hard to identify to identify cost savings. But, um, I think today you don't have the choice anymore [00:37:30] to not use it because other business will use it, and then you will start to lose market share because for the same price, another company is going to offer more value and maybe with less people. So that's, uh, where I see. And you can I mean, I read a lot of articles that people expect the value, but nobody has the ROI yet. I think maybe in marketing agencies it's easy because, you know, oh, I needed, uh, ten people, uh, this month for [00:38:00] ten clients. Now, next month, I only need five people because I write everything with ChatGPT. But I talk with marketing agencies, and they say 8,090% can be done, but you still need 10% and that only humans can can do if you want to stay at the top.

Isaac Heller: Yeah, definitely. We see that too. Nobody's going anywhere. I guess the, uh, the challenge becomes, how do I internally become more of an AI champion, right. Because if I can say that I can move faster or produce more value, that's [00:38:30] a harder business case to a CFO than hey, I can reduce headcount in my supply chain, or I can reduce the cost of something in my supply chain. I also love we have to. We have to be inspired by our friends and the French in France, who really know how to have good quality of life and work hard. So I always, I don't know if that's, uh, that's in the blood or just you, but, uh, you know, I think that's really good to have some of the work life balance. I mean, look, here in America, [00:39:00] you know, you have people that are spending 50, 60, 70 hours a week in audit or in their closed processes staying up late during quarterly closes. I mean, it really can be, I would say, a taxing, no pun intended at times. So perhaps I will give you a little bit more continuity or or smoothing out. Now, one thing that you mentioned was Python. I, I haven't heard this as much. Uh, you know, truly in my, my partner is our product [00:39:30] and engineering leader. We have many people in our company that use Python and React and some of the best code. What's what's interesting is there's a whole AI world around, uh, coding and engineering. But it's also interesting that you mentioned in the accounting world, having those Python skills might be increasingly valuable. How do you see that connected with AI and some of the things we've been talking about?

Nic Boucher: Yeah. So for me, 2023 was the year of ChatGPT [00:40:00] and generative AI. And now 2024, if you are in finance and you already master generative AI, you should go to Python. So Python is the new tool to learn, um, you know how pivot table, how, um, just mastering Excel was really needed in the last years. Now Python is becoming this new tool because Excel is not enough for a reasons why. At first, you cannot do complex visualizations within, [00:40:30] uh, Excel. Like try to do a heatmap, try to do, um, even uh, a map, like a geographical map is really, really hard. And you need a lot of time while in Python. You just need like a mini line of code, and then your data is represented in a heat map. Then second in Excel there is Power Query, but only a few people know it. But it's still really long. You need to click everywhere to automate your [00:41:00] data processing, your data crunching if you want to combine files, and with Python with just one line of code, you can combine, for example, 100 files together and takes one second. And now the beauty of it is that this line of code, you don't need to know how to code it yourself. You can just ask a generative AI tool to get the code and then the work is done. Then the third part is that for [00:41:30] complex financial analysis, you need to build a really complicated model in Excel while in Python. Again, if you use a coding with a generative AI, you can build a model really quickly. And the fourth one was actually we are all we all got promised to use AI for forecasting. But you cannot do that in Excel because Excel doesn't have a forecasting function. [00:42:00]

Nic Boucher: So Python has that. Python have has um algorithm for example like uh prophet, profit, which is an algorithm from Facebook that you can just download it for free and you can use this algorithm to forecast, for example, your seasonality of data, plus a trend on your revenue. And what you just need is historical data, and you apply the code on it, and it will give you the trend for [00:42:30] the next 12 months or three years. And this is really available now for everybody, but only if you accept to break the barrier of not being afraid by Python. And the good news is that it's actually now really easy because you don't need to install Python. You just need to use something, uh, like Google Colab, which is a place like Google Doc or Google Sheet where you just need to put the code inside. And second, you don't need to know how [00:43:00] to code. You ask generative AI with your own words how to do the code for Python to combine 100 files, and it will do it for you. So there is really no reason anymore. And we teach that like I have a tomorrow evening, a new training, a new cohort where we show with tech CFOs, with We take fractional cfo's. We take head of finance. We show them in ten minutes how they can use Python. And you should see their eyes opening and saying, wow, like, there [00:43:30] is so much I can do now and even myself. I never learned how to use Python, and every week I'm using it to combine files and to automate some of my processes.

Isaac Heller: That's amazing. Yeah, I never understood why they named it after a snake. You know, most people are afraid of big, uh, big snakes. But Python has become the universal, most, most common kind of, uh, code these days. And I also think it's cool that you have ChatGPT almost as your assistant to help you learn [00:44:00] Python, right? You don't have to be an expert overnight. Um, and you also don't need to learn all the deep nuances. You can almost have a copilot in in ChatGPT or whatever it is to help you learn some of those Python things. You also mentioned that there's CFOs and heads of finance that are that are learning Python. I think that's super important. I mean, I think there's there's some people out there, they're going to say, okay, well, I'm at this point in my career, you know, I managed a team or I've managed teams of teams and all [00:44:30] that. I'll let those guys figure out, you know, all the AI and all the Python and stuff like that. But I think even going to a class like Nicholas or any type of ongoing education can be very valuable because you want to be dangerous. You want to understand what they're thinking, you want to understand what they're learning.

Isaac Heller: You want to maybe maybe you'll even suggest, hey, have you thought about using a Python script for that forecast that you're trying to automate, but you don't just want to say it because you saw it on, you know, Nicholas LinkedIn like, you want to say it because [00:45:00] you felt it and you thought about it. And I think that's that's super important. Um, it feels a little bit in talking to you and our customers that there is there is a feeling of some level of renaissance in finance and accounting as it relates to the businesses. And, you know, sorry, the new types of tools, whether it's Python or AI. And my sense has always been, you know, in America, I don't know how it is in in France or Germany, but the nickname is bean [00:45:30] counters. Have you ever heard that one? Uh. For accountants? Yeah. Bean counters? That's not really a fair name, because it. The impression is, you know, someone who's only looking at their computer, looking at the spreadsheets or whatever it is, in my experience, growing up in finance teams, the financial leaders were the most quantitative and tech savvy people. Right? That's the other thing is we assume, because, you know, the old Oracle or SAP interface is so old or Excel [00:46:00] seems so rigid that it's not a tech savvy note.

Isaac Heller: No, listen, finance and accounting is the quantitative field. I mean, probably adopting Excel in the early days was one of the most advanced areas. So I just have this general impression. I've always felt like this that, um, financial leaders are early adopters. They are highly technical. They are quantitative. The only thing is, you know, they have a very high bar, like we have a very high bar for what needs to work. And we're okay taking bad UX, bad [00:46:30] UI, something that's old as long as it works. Right? And what does it mean to work? The numbers are right, right? The numbers are accurate. The numbers are reliable. That is the best feeling in the world and the worst feeling when there's one, you know, decimal error or whatever it is. So maybe just like in our last, um, kind of few minutes, I mean, this this has been like great talking to you, but talk a little bit about, uh, upskilling and then just mastering AI and why finance is the [00:47:00] right place or the right industry to to mastery AI. Give us some inspiration, you know, coming out of this and being more curious and, uh, advancing our, our domain.

Nic Boucher: Yeah, I really agree with what you just said. I think people in finance are too humble also because we don't see how much we have to deal every, every day. Uh, we need to master the ERP. We need to, uh, to know everybody [00:47:30] in all of the departments. We need to deal with a lot of changes. We are at the end of all of the processes, and it needs to work like the fingers work. And even if when the figure comes is not right, we need to make it right. So because of this, finance is actually the right place to be. Champion of using AI. I often use it as for me, my objective. I really want the finance teams to be the champion of AI in business because [00:48:00] AI is good with data, because it can process much more data than what we can use is. Ai is really good with forecasting. Ai is also really good by using I'll say like non structured data and to structure it. And for now like in finance that's the part that we do it with our brain. When you go and you talk with your colleagues from the business and you get to hear that there is a change in the contract, but you don't see that in the figures. That's where you use your brain [00:48:30] to make something qualitative and you connect with the figures. And for this, because we have the best knowledge of the business and how to translate it in figures, we should use AI to do that.

Nic Boucher: To do that at scales, because we are the best place in the business to do that. And I really believe we should stop being humble, but really like use all of our skills, which are not only been counting, but technology, knowledge of the [00:49:00] business and also we are really good, I think, at connecting the dots between all of the departments. And because of this, we should be the ones that are the first one to show the value of the of AI, and especially either in forecasting with machine learning or either if you look at the insights from the historical data to make some proposition for for the business. So to everybody listening to us, you you are already ahead. Normally [00:49:30] if you are listening today, because a lot of people choose not to listen on their extra time. So if you are here, that means you are ahead, but continue to be ahead by learning how to use AI for finance. And also once you learn, don't keep it for yourself, just share it around because we also need to grab everybody with us that don't have the chance or the appetite for curiosity, but they need to come with us and to continue to move finance towards the future.

Isaac Heller: I love it, I love it, and [00:50:00] I just I have to share. I can't hold it in. I was speaking with, uh, Chief accounting officer a couple of weeks ago. One of our customers, um, there are there are mid-cap, uh, public company and chief accounting officer. And he recently got a promotion in the past year. Um, he's now also the CIO. Right. So CIO and chief accounting officer, I haven't seen that before. I mean, it's very talented. And, you know, with his permission, we'll, we'll we'll do a we'll do another interview. But, um, he, [00:50:30] he implemented, uh, Julian. Right. And it was one of the only I use cases across the business. And now in 2024 and 2025, they've identified 25 use cases, not just in accounting, actually the lease or in accounting. And most of them are across the business and customer experience and supply chain and everything. So I thought that was a very, very interesting career path. I don't know if that's going to be an exception or something we see more often, but it really reminds me of what you're saying, which [00:51:00] is the finance and, you know, accounting team can be the next champion of AI and technology. Um, last thought, what's next for you? Are you, uh, what are you most excited about? Is it your course on Python? Is it a new roadmap that's coming out? Give us some preview of what we should look out for next.

Nic Boucher: Well, I'm the most excited with is radio I. Finance club, because that's the place where every week we bring a new piece of content. For example, this week [00:51:30] we released a comparison from, uh, made so written by the chief data scientist. Of a bank who showed us how to, um, the comparison between the open LMS and the property. Um, so think about Mistral, for example, for open arms and proprietary. So OpenAI and when to use which one based on your business based on your constraints. This week I'm going to release [00:52:00] a mini video course on how to use ChatGPT for all the data analysis part, so or code interpreter. And also soon I'm going to release a mini course on code 3.5. Because with artifacts there is a lot of new cool things that code can do. And we have a master class in one week where we are going to talk about how to build your own AI roadmap in finance. So that's for me, the [00:52:30] really the right now, the what makes me the most, uh, the busiest and, uh, excited that more people will join because there is so much information inside now, after seven months that the club exists and you you can imagine, like every week, a new piece of content times seven months. Now, it's already a lot to learn.

Isaac Heller: I love it. Awesome. Okay, good. So encouraging everyone to check it out. Nicolas Bouchet um, nearly a million followers on LinkedIn and [00:53:00] will post, um, and not just a face of finance, but someone who backs it up and I think really has some good original insights into this industry. So as always, Nick, really appreciate your time today.

Nic Boucher: Thank you, Zach. And uh, good day to everybody listening to us.