AI in accounting: Demystifying the present and shaping the future

Hi, everybody. We are just waiting a couple of minutes as everybody gets joined to the webinar. We're excited that you're here. So we'll, we'll give it a minute or two and then we'll get started.

Hello, everybody.

Hello.

And while you're joining, if you'd like to say hello in chat and let us know where you're joining from, that would be amazing. And just a reminder too, if you have any questions as we're chatting, feel free to use that chat as well.

So people start answering.

Yeah. Austin. Dubai. Excellent.

Switzerland.

Singapore. I was just in Singapore about a month and a half ago. For the first time ever is Singapore. I, I don't know. You haven't been. It's an incredible city.

Yeah, I've lived and worked two years in Singapore.

Oh, amazing. I was there for their annual, like their big, you know, Independence Day. It is incredible. Like, they had jets and like the shows, like it was, yeah, great time to be there.

Awesome. All right, well, as people are joining, I think we can get started, so. Hello, everybody. Welcome to our webinar. AI and Accounting Demystifying the Present and Shaping the Future. My name is Mandy Schnurl. I am the vice president of marketing here at Net Gain and I'll be your moderator today. Thank you again for joining us. We're very excited to explore today how artificial intelligence is, is transforming the accounting industry, helping teams to enhance data accuracy, automate repetitive tasks, and even tackle complex challenges like financial close and compliance. And we are very fortunate today to have two industry experts here with us. First, we've got Adam Riches. He is a cpa, CEO and founder of Net Gain. His extensive experience in accounting technology has helped to drive innovation in automation solutions. And welcome to Nicholas Boucher, founder of, of AI Finance Club and thought leader in AI driven finance. Nicholas brings insights into AI's potential to revolutionize the industry as well as the practical applications we can expect in the near future. Would you all like to say hi before we move on?

Yes, everybody. Yes. Hello. I'm, I'm in Denver. So we're seeing all the. Everybody drop their locations in service. We've got Dubai, the Netherlands, New York, Chicago, Somalia. I think I saw one from Somalia. And so we've got. This is great to see people from all over. So I'm in Denver and then. Yeah, Nicholas.

Yeah. So I'm French. Like, everybody can hear with my accent. Impossible to hide it. But I'm based in Germany and excited to talk about you because while preparing this webinar, I discovered How Adam is like one of the most knowledgeable person I know about accounting and all of you know about the history of accounting and you're really passionate about it. Where does it come from?

I Yes, I love like I think we've got a very, very rich history and especially like the modern accounting, like the double entry Italian method of accounting is very exciting. I even have, I've got, I've got my. This is a book that was written published in England in the mid-1700s about the Italian method of accounting. Like I love yes. Everything about our profession. So and especially today, like this is a great time to be alive. If we look at the, you know, the compounding information that we have in our profession. It's very exciting. So excited to be on here with you Nicholas, and on this webinar.

Very excited to have you both. Should be a very interesting discussion. So what will we discuss today? Today we'll talk about the current role of AI in accounting. So how is AI helping to automate tasks like bookkeeping and month end close? We'll also talk about the future innovations in AI where we'll explore emerging trends such as predictive analytics and open banking. We'll also discuss real world applications where Nicholas and Adam can actually share some examples of AI tools that are solving complex accounting challenges. And then we'll talk about how does this apply as we prepare for the future of AI and accounting? What will the data challenges look like? How can we prepare for it, and the importance of clean structured data so that your AI implementations can be successful. And before we dig into our discussion, just a couple of housekeeping items. We'll have a Q and A session at the end of our discussion today. So feel free anytime to drop your questions into the Q and A box or to the chat. And also we are recording the session. So after the webinar is over, we will be emailing out a recording. So you'll have that for your records. You can share it if you like, but you'll have that available to you. And so as we get started, we can start our discussion here. And first things first, I'd like to kick us off with, with a poll. So as we get started, let me just get that going for us if I can find it. Here we go. All right, so as we get started, we'd love to know how you are using AI in your organization today. So you know, I think we've all started to dabble with AI to maybe adopt it officially. But how is your organization tackling this? Are they adopting it across Your company, do they have restrictions? So we thought this would be kind of interesting to see as you are all responding to it. Nicholas, what have you seen among your clients as far as organizational adoption of AI and finance? As we're waiting for folks to respond.

Yeah, so you really have four types of organizations. You have. The first one is more small companies that are already tech oriented, they embrace AI. Like everybody needs to use AI. Like the management is behind and push it. They also organize training for AI. So with generative AI, but also with machine learning and forecasting, then you have companies that are more, I would say like bigger companies, but still in the tech industry. So I helped a company called Couchbase or another poshmark in the US you should know them. And they have contracts with ChatGPT or with Azure for Copilot or with Gemini from Google. And there you have the use of AI, but with a contract with big LLMs. So employees have to use that. And then they also encourage, but that's how they organize to make sure that everybody is using it, but in a framework. And then you have the third type, it's more industrial companies where the management is still hesitant about where to go because there is a lot of rumors about what AI can do with the data. And so they take their time. They do like small projects. Often the IT and the developing team, they have access, but other teams, they need to ask access. So that starts slowly. And then you have the fourth type of company which are more like for example finance institutions, an old finance one where there is a big pressure from outside from the regulator. Especially in Europe we have the EU AI act where people should not use AI because it's too risky. Those ones are also slowly starting to use generative AI, but they really take their time because they have so much pressure from the regulatory environment that it's hard for them to start using generative AI.

Absolutely. Okay, let me go ahead and end the poll so we can see our results. So pretty interesting. It seems like about only 9% of our attendees said that they are not allowed to use AI at all in their organization. But overwhelmingly it seems like most folks are permitted to use it, maybe with some restrictions. Are you surprised by that?

Well, your assessment, Nicholas, feels like like exactly what we're kind of seeing. And so from our perspective here at Net Gain, we deal with a lot of NetSuite customers and QuickBooks customers. And so we're, we're kind of mid market and down. And I think companies in that space tend to be feeling so much pressure that any Relief is welcome. And so we don't see a lot of concerns around the adoption. But what we see on the, in the tech side for tech companies like that are building solutions around AI. The bigger organizations tend to have a layer where legal kind of sits in between the dev. Org and releasing product. And that's created some really big challenges, I think, for the, for the bigger organizations that want that. You know, they're all excited and they're all talking about AI, but there are, I think, some real restrictions in what they're able to actually release and launch because of that legal layer that sits there. And there's, you know, legal's obviously a very concerned party, I think with AI and, and so the bigger companies tend to be a lot slower in innovating with AI than a lot of the smaller startups that, that we're familiar with and that. And that you talk a lot about.

Yeah, very, very true. It's very interesting how things are kind of evolving as we go pretty rapidly too, which segues nicely into our first topic. So Adam, could you start off by sharing your PERS perspective on how AI is impacting accounting today and where you see the biggest benefits.

So I love, I always like to take like a historical view of this and, and I think the last major, major innovation that we've had in our industry, that's why it's so exciting to see this come out now, was excel in the 1970s. I think that really changed the way that we work as accountants and really made us more efficient and able to, to take a different role than I think our profession used to have. And so this one is exciting now. If we look back, I think LLMs, like the way we kind of think about AI, it came out about two years ago, a little over two years ago, and there's been a lot of excitement around it. You know, how this is going to change. This is going to fix like the shortage of accountants. Like this is going to turn accountants into like very value add employees instead of doing a lot of the busy work that we've been stuck with. It's so time consuming. So with all the excitement over the course of two years it has been, you know, and there's a lot of big companies, tech companies, talking about AI and how they want, you know, this is, this is one of those major pillars. It's been frustrating to see the practical examples come out. I think we've been slower to like really get, you know, and execute with a lot of the practical ways that we can leverage this again, that's why I'm excited to be talking to Nicholas. I think that's one of the core pieces of AI Finance Club and what you're doing. And it's also one of the things that we try, you know, we're aggressively trying to be on the edge of, like, how can we leverage this technology to actually move us in the right direction? And so. So I don't think it's been the white knight that we've all kind of envisioned it would be two years ago today. I think it absolutely has the potential, but I think that that is largely dependent upon us innovating and figuring out how. How we can actually use this new technology.

Yeah, I think that's really spot on. And, Adam, as accounting standards become more complex and regulatory compliance just continues to evolve and become more of a burden for finance teams, do you think that AI could be a possible solution for businesses to help meet those regulatory requirements in an easier way?

I think there's a few pieces where it can go take us a long ways. I think it'll help advance across the board. I think for statutory, like, for tax reporting, I think there's a lot of things that we can do to have that help us take us very far down the path. I think for financial reporting and for these technical accounting topics, I think it can do a really good job of serving up the information in the right way. But I think we're still going to need the expertise of, like, CPAs and chartered accountants to come in and really be that final checkpoint before we start processing journal entries. I think there's always going to be that layer that will need that layer of expertise on the technical reporting side because there's just so many nuances across industries, across businesses, and like these internal accounting policies that companies have, I just, I think it's so unique, but I think it can take us a long way so that you're spending less time pulling data out of a contract and more time focused on what the core issues are around, you know, like. Like some of these accounting treatments.

Yeah, absolutely. Thank you. Thank you for those insights, Adam. Nicholas, could you give us a glimpse of what you think might be next for AI and finance over the next couple of years? And what do you think the. The power to transform the space could be?

Yes. So if I want to share something that I prepared actually where I think there is also the adoption today is not only about how is the future, but also your size. Because if you think about companies like, let's say, Google, Amazon, even the big banks, they already used AI since a long time, so they are already in advance. But you need to think about AI first for your team. So the first part is for smaller companies or for smart team, the first thing you need to do is to use the tools that you don't need six months to roll out or you don't need to buy somewhere else. You have ChatGPT, Copilot, Gemini, they are available for everybody. You can either have it enterprise. So like this, you make sure that everybody stays confidential. But also for smaller companies, employee can use it if they know how to use it without using confidential information. And that's a great way to push acceptance, to push adoption, because those tools are there and you can already make people work with AI without having to wait a finance transformation. And the second part is the process that actually us in finance, since already like 10 years we are using AI and we don't think about that. But the second process here, the second line is digitalizing invoicing, invoices and also process it. And now the big development since 1, 2 years is on top of having AI recognizing the invoice. Now you have also AI who understands. And because AI understands AI can map the invoice to your ERP and there you have a lot of also saving because you don't need humans anymore to be there to map the recognition with the server. And the last part, if you are a small team, you should start using Python because now you don't need to code it, you make AI so generative AI code it for you. And Python is really the tool that can allow you to start automating because it can write Excel files, it can change them, it can create graphs and also you can start also pull data from everywhere and even create reporting with this. So if you are a small team, start there. If you are a bigger team or a bigger company, you can be a bit more ambitious and you have a bit more money and also what is going to come in the future. And there that's where you are going to start investing in AI native tools. But today there is not one AI tool that covers all of the finance processes. That's really important to know that. So what you will do is you will look at one process. So let's imagine that you want to look at your revenue recognition. Well, you can use an AI tool like there is one called Trulion, where it recognizes on the contract all of the new recognition criteria, reads it for you and books it in your system. And then the second part is you need to start forecasting also with AI and for this Python you can use Python and machine learning algorithm. And the last part is in all of your automation that you already have now in all of the RPA processes, you could call the API of those OpenAI tools or Gemini or Amazon and you can generate text based on the event. So you can like this for example, tailor your reminder if you want to have it tailored to your process. And the last one does more for bigger companies. So that's what is going to come for big ones. They have their own business model, data model, but they will also have their own forecasting models like Microsoft and Coca Cola. They already have that in place. So I expect bigger companies having that. And also we are going to see more and more AI chatbots either today just to answer, but also soon AI agents. And we here by with Copilot from Microsoft that is already as From, I think the 24th of November already going to come. And AI agents next year is going to be really real and not anymore buzzword. So that's what's my big expectation is the move from having a chatbot to have an agent that can read your system, but also write in your systems and do tasks that are really repetitive and that you can delegate to this agent.

That's amazing. Yes, thank you. I love how you broke that out into the different company sizes too. Nicholas, if you had a crystal ball in kgage, even further into the future, maybe 10 years from now, do you see any emerging technologies coming into play that you're really excited about? And how do you think some of those technologies might reshape the role of accountants and bookkeepers?

So I think continuing on the AI agents, what we'll have is as a finance manager, it's not anymore that you will have to manage teams of humans. You will also have to manage teams of humans and agents. So robots and the way you manage it is actually the same way than a human. First, you start by recruiting. So you can buy agents that exist or you can also build them. You also have students that come and you shape them in the way that you have the good culture of the company. You can have the same with agents. You can start building them internally and then you review every two months, three months, you review if the process is still adapted. So if either your human capital is still adapted or your agent and you keep training them, you keep coaching, training to make sure that the agent and the humans are working well. And every year also you see if you need to change these agents to change the scope. And it's not anymore that Those agents and AI tools are for it. It's every manager will have their own agents will manage them, but we'll also ask the team to manage them. And that's, I think, a big change. Because think about it, who wants today to record thousands and thousands of receipts from restaurants receipts or travel receipts? Nobody wants to do that. But that we can delegate to AI agents. Who wants to spend a lot of time on the phone calling people, why you didn't pay, why you didn't pay? A voice agent can do that as well. What we should do more is go outside of our finance room, go to the shop and ask our partners, like, why are we having lower sales? Why is there a machine that is not working anymore? How much is going to cost? What should we do? What do you need from us that we unblock money that we can have faster sales? And that's what finance needs to do more and more. But to do that, you need also to have your technological skills to manage these agents.

That's awesome.

I agree. What I think is exciting, like RPA has been around for a while, but you start to take some of these existing technologies and then you combine. It's like compounding technology and AI on top of some of these other technologies can be incredibly powerful. And that's where I think our understanding of technology and how to build these and create these agents. To your point, Nicholas, I think is going to be a necessary skill for our profession so that we can focus on those things that really matter in the business. I think, Mandy, one of the technologies that has been largely overlooked over the last decade that I get very excited about is there's this open banking revolution. So our data at the bank, if you log into your bank statement, like I'm sure all of you do, to get the books closed, right, you get your date, you get weird memo field that is mostly understandable. And then you get the amount that was added or removed from your account. So you get some very basic data. But I think we all understand there's huge amounts of metadata that sits behind those transactions. And the bank historically has kind of held onto that information and has only released the basic data for us to see on our bank statements. But with this open banking revolution, there's areas, there's legislation around the world that's now talking about forcing banks to kind of release this information. And it's already happened in Australia, New Zealand. Like, they've got requirements where they're. They're required, you know, with the consent of the company, like, to release this open banking information to the business so that those businesses can then go deeper on what, you know, what's happening behind this, behind the curtain for these different transactions. So some of the information you get with open banking is like you get the geolocation, you get a timestamp, you get, you get merchant information, and sometimes it's even enriched by AI. So phoenicity here in the US enriches some of that data so that you get better data points around it. So for companies like, if we can take the open banking data that comes in and apply AI on top of that information, we can now start to look at like fraud detection or anomaly detection or even pattern recognition as it comes in directly from the bank. And so Instead of taking three data fields, we can now take 30 and use those 30 to really. And even then like matching data on your bank recs and things like that. So I think, you know, RPA is one powerful, like, combo, like thing that you can layer AI into, but then open banking, and layering AI into open banking, I think is another way that we can really get high quality, like really that bookkeeping, the recording function at the ground level, getting all of that information into your accounting system at a very high level of quality with very little human touch, I think is one of those exciting points that I know I could talk for days about.

So anyway, yeah, that would be exciting. Go ahead, Nicholas.

And on top, I think until now, we didn't have really the tools to process this data that is available now. We have these tools, but we lack those data. So having open banking available from everybody with just like API call is what you need to start enriching our analysis. Stop having to reconcile everything manually, but having enough elements to have automated reconciliation also to plan your cash better, because with more and better bank information, you can plan so much better your cash because you can recognize those patterns from clients or from vendors about where the cash comes in, where the cash goes out. Or you can also detect some fraud that happened because the last, I would say, like safeguard is the bank account. And if you have more information about what happened, then your fraud detection can be better.

Exactly. We can see if Mandy's spending on luxury items or gambling on her on her company card, that'll be easy to detect.

I would never. All right, switching gears a little bit. I know there's this growing concern. There's fewer and fewer people entering the accounting space. And because of that, it's becoming more competitive. There's an increasing cost of bringing accountants onto the team. How do you think the AI could address some of these challenges. And I'll open this up to both of you.

I think we went through this, we went through this with spreadsheets. We had armies of accountants. It's interesting to see, you know, before, well, there was the systematizing of accounting with the mainframe computers in the like 50s and 60s. And then the 70s brought the spreadsheet. And what happened before that is you have to have an accounting warehouse. You had to have an army of people tallying data at the end of the month to close the books. And so a lot of accountants, their function was Simply like with a 10 key running numbers, right. And if there was an error down or upstream, then they had to rerun the numbers again. Right. And it was a very time intensive process. The like the whole function changed with spreadsheets. And I think, you know, with that one account was able to do the jobs of like many people, you know, what it took an army people to do before with the spreadsheet. And I think we're about to see the same thing. I think as we, you know, level up like our technical expertise, we become experts on AI, like we're going to be able to figure out how to do the work of 10 or 20 or 30 people. And that just increases the value of the role that much more. So I think, you know, so what happened in the 70s is the value of accountants went up and the pay went up accordingly. And I think the same thing, this will drive an increase of wages for those that are early adopters of this and on the bleeding edge. And I think the wages, I think it'll follow as we're able to take on that much more. So that's my thought and I think it'll resolve itself. I think that'll increase the demand and we'll see this resurgence in accounting here. That's, that's what I, how I feel. I don't know. Nicholas.

Now what you hear everywhere and from us, Europe, Asia is that it's actually difficult to find accountants. It's actually difficult also to find people that are flexible. A lot of people are going into retirement and younger generation took less master in accounting or like accounting education. So I hear already like some CFOs, finance managers that have difficulties finding the right stuff. So it's not that AI is replacing jobs of existing people having the jobs. It's more that because companies don't find the right talents then they have to go also to the technology to replace the jobs that are not fulfilled. And also I think since COVID There is really a big change about the importance of what we do. I noticed that everywhere people are re questioning is it normal that I spend 10 hours a day in my office just behind a screen and just typing something? Should I not do something better? And that's where technology can help you. Because in finance we are actually lucky. We have access to a lot of departments, we have access to a lot of data, we have direct access to the management. So we should use technology just to make our job better, faster and have more fun and have more impact because we could really impact. And just to illustrate what can AI do because I think it's important to show to people I had a client and we were doing a training and this client is getting acquired by IBM and so they will have to go through accounting policies compared to see what are the difference between the accounting policies of IBM and this client is Ashicorp. So imagine the work of doing that for those big companies. Well, I told them I can do the work for you in two minutes if you want. And they were looking at me. What I did is I went into the new model of ChatGPT and I just say that there is this acquisition between IBM and Ashicorp. And then what I did is I pulled out the information from the financial statements that are public. So it's not confidential information. I just copied this information. So the accounting policies of Ashicorp and the accounting policies of IBM. So it took me more time to find the accounting policies than you will see do the analysis. So I copied that and then when it was copied I went back into ChatGPT and paste that and ChatGPT started analyzing one by one all of the policies. And you can see that really quickly it identified all of the gaps and also the similarities between the two. You have to think how many people will have worked on that or how long a team will have worked to do this analysis. And it's not accelerated. It was really that fast. And then after I just asked to show it in a different way and the outcome was this overall recommendation first is not the same fiscal year, the end for IBM is 31 December and Ashicorp was 1-1-31st. Then you had of course one part I'd really like is the deferred contract acquisition cost but also capitalization of software. Those companies, they have huge servers and huge also ip. So I mean Adam, you as a accounting fanatic, you will love to see these things and to have to work on that. And what is interesting here is the hard work of really like Going through point by point, you can delegate to AI, but the intelligent work of understanding. Okay, I have a gap. How do we deal with that? What do we do? Like do we have to go full on follow what IBM does or do we have to negotiate with them because ours maybe make more sense for our business to continue to use it? And that's where human you will have more time to do it, but also more pressure from outside because everybody will know that you can do the analysis in two minutes. But what they expect from you is that you spend more time delivering more value because now you have the time for it. And so if in finance it doesn't come from us that we increase the value we bring, it will come from outside because everybody knows that we have more means. And if you go today to boards or to presentation, people want to see amazing graphs. They want to see amazing presentation, a dynamic dashboard, a live data. Like 10 years ago you just showed a graph, you were the star of the meeting and just demand increase on finance. And that's what is going to happen also for us.

Yes, I love your example, Nicholas. And yes, as an accounting nerd, I was digging into some of the policies, but it's an incredible, I think it's that idea that having it as your co pilot is always an incredible idea to accelerate the work. There, there was, you know, some chat about, you know, what about hallucinations? Like, can you trust any of that analysis that was done? I do believe, like that's, that's an amazing, amazing first, very detailed pass. And as you go in and start to execute and start to merge these together. Yes. It requires like a deeper look and, and with prompting and with reviewing the actual documents, like you can get to the answer a hundred times quicker with or without hallucinations. That may happen.

Right.

If you're, if you're an expert on these matters and you dive into the material, you'll see it if there's some major gaps. But I do think hallucinations have been kind of overplayed, especially in topics that you're knowledgeable and expert in. I think hallucinations are threatening.

And if we take this example, you have two ways of dealing with this. Like now I could just go and give that to the CFO and I will have a lot of risk that maybe something is wrong. Or I could take another hour, read quickly and check if, yes, what is written is correct or not, and make also my own interpretation because I need to own it. And it's exactly the same way that if I ask my staff to work on it. I get the result and I go to the CFO without looking at it, or I take another hour and I review if the work of my staff is good. So it doesn't matter if you delegate it to technology or to humans, you still need to own the result and you need to review. And the way you get the quality of your team or the technology is the way you coach it, the way you parameter it, the way you instruct it. And so you have a lot of importance about and a lot of a big role to play about how you use these tools the same way than how you coach your team and how you review your team. So that's just like. There is no like difference between the two. It's the same at the end. You are managing those resources and use it for your own. But once you use it, you own it and everything doesn't come everything that comes with it. If it's great, it's your work. If it's not great, then you cannot say, well, it's not me, it's AI who did it. So that's normal. That is better.

I agree. You got to own the. Yeah, exactly. You got to own the product that you're putting forward. So absolutely. Yeah.

I think that's why human oversight and the knowledge and skills that accountants have will not easily be replaced fully by AI. Right. You still need that knowledge, you still need that oversight and validation, especially when it comes to training those models. And Adam, I know that you had some examples for maybe some more day to day practices for accountants and how they could incorporate AI in just time savings and efficiency.

Yeah, yeah, yeah. I saw there's a comment for somebody with a company under 100 million. Right. Like the public filings, like is less relevant for maybe a small business. But I think the principle stays the same.

Right.

Like having it as your copilot as you run into different topics or challenges can be super helpful. And from our perspective, we're always trying to find ways to embed this in technology. So I've got a couple of examples that we're working on. One Nicholas mentioned earlier. Right. So we have a lease accounting tool and you can take your 60 page lease agreement and load it in and we will analyze that 60 page document and pull out the relevant data for your lease record and populate the lease record. And of course like everything that still needs to be reviewed, you still need to own kind of the data going in and the treatment for that lease. But we really serve up the data in a way that makes it so instead of spending hours on the document, you get to spend your time making the right decisions on how that lease is treated. So that was one example. But another piece that we're working on is we have a bank reconciliation tool embedded inside of NetSuite. And so we pull in open banking data. And one of the exciting things we're working on there, we have a list of rules to match transactions against general ledger transactions, bank transactions against general ledger transactions. And from like our perspective, like the rules are always the highest priority. If you've got a rule that a certain vendor should be treated a certain way, that's the easiest and best way to map these transactions into your general ledger or automate the creation of a transaction. But if there's gray area and if the rules aren't perfect and they don't apply to these transactions coming in, then there's a second tier. And that's where we're working on a model to build out where the AI will auto code or auto like kind of predict what those transactions, where those transactions should be coded or where they should be matched to in the system, as well as looking at like potential rules that could be created so that you can avoid the confusion in the future. And so these are some of the pieces that I'm excited that we're working on. But there's a couple of other practical areas that I wanted to highlight here. So let me share my screen really quick. Let's see right here. The first one that I wanted to go through is what Nicholas talked about. And I think we see this a lot. It's kind of that OCR like the recognition of data in an invoice that's coming in. So instead of keying it in manually, it can be automated. So we've got a fixed asset to tool. And if we, you know, go in, we can upload a vendor bill for some assets that we purchased and it will read this vendor bill and it will take a look at, you know, the, the AI will analyze this. You just drag and drop this in and it will look at all of the, the details down here around this asset purchase and create the system transaction which is on the left hand side of the screen here. And from there we can easily take this. It looks like this is for an excavator that we've purchased with a bunch of different components behind it. We can take these and create an asset. And this is a compound asset with all these different components. But we've created it and rolled it up under one single asset without losing the Detail, So just that easy, like we can drag and drop and bring it in. And I think these are some of the things, Nicholas, that have been around for a while. I think they're becoming far more elegant now with the latest and greatest AI tech that's out there. But these are some of the things you've probably already seen. So we've got this, we've got our lease accounting tool as well. But one of the things that has really got me excited is on the flux analysis. So we have a CLOSE tool again embedded inside. And that suite, we'll be coming out with it outside of netsuite as well. But right now our solution for CLOSE is inside of NetSuite, but we have a flux analysis. And as a reminder to everybody what a flux analysis is, it's at the end, when you close the period, you're looking at your prior period results against your current period results. And any accounts that have significant fluctuations as a percentage or as a dollar amount could trigger requiring an explanation behind it. And the idea is this is an overarching control that will help you detect any errors that may have made it all the way through to your face financials. And so what the benefit we have is by being embedded in NetSuite, we've got all of the underlying transactional data that we can look at. So what we're able to do is anonymize that data and throw it out into one of these LLMs to produce an explanation of what that flux should be. So instead of the old way of doing this, looking at taking a data dump of the prior period transactions, a data dump of the current period, and then using some pivot charts and magic to try to identify where those big differences might be, we can throw that data in and have, and have AI generate the difference for us. So I'm going to run that right now. And you can see here, this is some of our data. So I've got the other information grayed out. We obviously use this internally as well. But you can see In June for Q2, we had 150,000 that was spent on marketing expenses. And then in Q3, that went up to 365,000. So we had a $200,000 delta between those two quarters. So this is one of those areas that we really need to dig in and investigate what the differences are. So again, using AI, we're able to generate an explanation of the differences based on all the underlying transactions. And the memo fields and the in the vendor information that sits in there, we're able to AI is able to take a first pass at this. So to your point, Nicholas, the controller, whoever's responsible here has to own the explanation that goes in. So this may or may not be a hundred percent accurate. So. But looking at it, I can see there's some things in here. I can see an Oracle expense. So what we did is we sponsored Suite World, which was in Las Vegas in September. So that came through in Q3 and that was one of the big expenses, one of the major drivers for the increase quarter over quarter in spend here. Another area is, so we rent out the High Roller, which is like the London Eye in London there, if any of you guys have seen that. So we rent it out and we have an event where we invite our customers and partners to come and join us on the High Roller for a couple hours. So that was another big expense that we had in Q3. So what I can do, knowing what I know about Q3 for NetGain marketing, is I can then add context to this explanation. So I can tell, you know, the AI engine what we did in Q3 and then have it add that in to the variance explanation that it's come up with. So now you can see it's going to rewrite that variance based on the new information that I've given it. And it's, it's a lot more accurate and close to exactly what caused that, you know, the drivers that caused that variance quarter over quarter. So this is one of those areas that hopefully can remove a lot of the burden, you know, that initial digging and busy work to try to find out where these differences are. And we can very quickly come up with an answer that can be enhanced to be the right answer. I can take this one verbatim and use this, or I can take this as a starting point, add in and busy work to try to find out where these differences are. And we can very quickly come up with, with an answer that can be enhanced to, you know, to be the right answer. I can take this one verbatim and use this, or I can take this as a starting point, add in some additional color and then, you know, finalize my variance explanation there. So we're taking one of these, you know, processes that seems to take forever right at the critical part of the close and accelerating that and actually adding a whole layer of accuracy. For anybody that's done it, I know I've done many, many of these. It's typically a roll forward and you kind of adjust whatever was explained last, last quarter. So we actually, I think, I think we get a much more clearer picture of what those variances are too, by, by using AI here. So these are a couple of the ways we're leveraging it. Again, we're, we're dabbling in a lot of areas to try to embed this into as many of our tools as we can. But here's just a few examples of what we're using.

So that's awesome. Adam, thank you for sharing that. And Anthony in the chat actually brought something up that was pretty interesting. He said hallucin is hallucinations only occur when the data being analyzed isn't validated and verified. So I think we all know, you know, AI is kind of only as good as the data that you have. And if it's garbage data in, it's going to be garbage output. Right. So knowing that many teams are struggling with poor financial data quality, what do you think some of the best practices are to ensure that AI can actually deliver accurate results? Nicholas, I don't know if you want to take that one first.

So here I have a bit of a contrarian opinion because I hear a lot of people saying, oh, you cannot start using AI if your data is not clean. Well, you can actually use AI to clean your data. You can use AI to enrich the data. And let me tell you how. First, if I look at one of my previous job, I was responsible for a factory of factory that builds these micro waves tubes in satellites. Each tube is the value of one car. And we were producing like I think 1,000 tubes per year. And so you have like 20, 30 contracts of these satellites being sent. And in each of these contracts we are seeing how many tubes we sold and which tubes and how much and also when we are going to get paid, when they are going to get delivered. These contracts were in paper, sometimes after PDF and it was somebody reading the contract and entering that manually in an Excel file. And then somebody else also took the PDF and was just focused on the delivery date. So again, Wen and Kirin that in their system somebody else again was just interested in the payment terms. Again, all of these three people were just typing in with the risk that they might type something different with the risk that somebody is sick. And so everybody waiting for them is also waiting for the information. So we don't have data or the quality of data is bad. Now you replace those data input with the AI tool that scans the contract and enters it in a flat table or directly in one of the operating system. Now first you have the data faster, the data is more accurate and the data is structured so all of the company can use it straight away. That's how you can first use AI to get better data, more data. Also because you might get more information captured than what you did before when you had to do it manually. Then once you have that, you could check in all of this list. Maybe something is missing or something is not correctly inputted. Well, you can run with AI, something called, for example, fuzzy matching. Fuzzy matching is. It recognizes if a word really is really close to another one. So, for example, if your company is called NetGain like you, maybe somebody entered Net Gain plus and actually it's the same company. So you can say, okay, I will remove the Net Gain plus because it has to be Net Gain like this. I don't have two companies. I have only one. And so you can use data to clean the. You can use AI to clean the data and then after have better data. And that's also how you can categorize. You could have so many general ledgers and saying, okay, let's classify that in several groups. And instead of having a human having to read all of the general ledger or all of the invoices, you can categorize them and do better grouping, better categorization to help on the data analysis. So, yes, you need better data, but don't wait. You can use AI to improve this data.

Yeah, it's interesting. I think there's a million reasons why AI is horrible and bad. I think it's like the Internet in the, in the late 90s, right? Like there was there, there's, there's, there's reasons to see issues with it. I don't know that many of them matter. Right. Like, I think we're migrating quickly in this direction, and it's figuring out not. Not what's wrong with it. It's figuring out how to leverage it and make it work best for your application. So in an ideal world, I think software vendors, you know, or Nicholas, like, we serve up these solutions that are on a silver platter. And I think in an ideal world, you just take those solutions and run with it. But we're so early on and we're moving so quickly that I think, and, you know, companies are so different in how they can apply and use this. I think it does, like, the burden, like, unfortunately does rest on the practitioners and figuring out the best ways to leverage this.

Yeah, absolutely. And one final question based off of that response, Adam. You know, as I think AI does continue to immerse itself throughout all of our daily lives, our work practices how do you see the role of accountants just evolving in the future as we adopt this more internally?

I love this question. It's one that I've thought about a lot and I'm curious Nicholas's take on this too. I believe that the ingestion of transactions will become, you know, like the manual entry of transactions will be a thing of the past. I believe that we will be on the front end. We'll be looking at anomaly detection only. So 98 of the transactions will fly through. They'll either be automatched or they'll be tagged, you know, or AI will find, you know, you know, where that needs to sit with, you know, 98 level of confidence. And then we sit and look at just the exception. So if there was a $50,000 payment out of our account that has no support in the, in the accounting system and we have no idea what that is, I think that's what, you know, the accountant's job will be. Instead of entering a thousand transactions in a month, they'll need to look and investigate three and, and, and even then I think there's a world where we could serve up even more data around that. There may be something in our CRM system that supports it. There might be an email that, that somebody on the procurement team sent out that kind of supports that $50,000 transaction. And, and I think it could even bring some pieces of the puzzle together, you know, on your plate so that it's easier to even resolve some of these, these, these issues that come up from time to time, right, where we have no support around a payment or an expenses come through. So, so I think we'll be, will be managing by exception. And then I think on the other side, I think we'll still need to be required to, to make some of these technical accounting decisions. We'll be able to manage like those a lot more easily because the data again will be served up to us the right way. And I think we'll, we'll have more, we focused a lot less on month end close. I think we'll have real time financials like I think with, with the, the integration of like open banking, leveraging, AI pushing this through the system. We're going to have a daily task where you show up daily and clear your queue of, of these anomalies and things that need to be whether it's a technical accounting issue or whether it's a rogue transaction, like, like we'll be doing this on a daily basis and we'll have much more visibility in a real time manner into our financials. And that will enable us to partner even deeper with the business and be less focused on the number crunching and more focused on how to, how to add value to the business. And at a macro level I think there's a lot of companies that struggle with data like slow closes and even understanding what's in your bank account or not even the bank account, but your real cash balance, the combo, your bank and your GL transactions and looking at what you really have access to for capital. And I think this causes a lot of inefficiencies in business in general. And if we can make companies 2% more efficient or 3% more efficient, the compounding impact of that and if we can get data in front of the executives a little more quickly, I think that has a major, major impact on how businesses are run. Historically the best businesses, the biggest businesses have been laser focused on their accounting and I think delivering that enterprise quality of accounting to small businesses and medium businesses will have a major, major impact on us over the long term. And I think I'm talking more the five to ten year range. I think now in five years I think we've got, there's, there's a lot there, we're dealing with a lot and we're moving 100 miles an hour to try to figure out the absolute best ways to leverage AI. But I think it's that 5 to 10 year range where we're going to see some compounding impacts of this new technology. So again I'm very, very excited. I'm very optimistic about the future here. So sorry with that I'll turn it over to you. I'm excited to hear your take on this.

Yeah, that's my dream really. And, and I think, and I hope, and that's my goal is that we are, we are gearing towards having super accountant where we are not compromised anymore. Where oh this, this person does only accounts receivable, this person does only accounts.

Payable this daily basis and we'll have much more visibility in a real time manner into our financials and that will enable us to partner even deeper with the business and be less focused on the number one.

Less focus does only data entry. Like we should not have that anymore and people don't want that anyway. If I take one example, I had a team where we had 10 people, three people went to retirement within one month and I just took over the team and we had to re challenge the way we were working instead of.

Crunching and more focused on how to, how to add value to the business. And at a macro level, I think there's a lot of companies that struggle.

Having people having just a focus on accounts payable. Then one month after she was also helping on the closing. And for her at the beginning there was a bit of doubt, a bit of oh, how am I going to deal? That I only did the last 15 years accounts payable. But if you take the time and if you show how to also use technology to have more time during the day to do this additional task, then this person just was so happy to do something new. To also understand better why she was doing the task from before because she could see the end by seeing the closing. And I think we should really go towards that. Go to the fact that we have super accountant that outside of accounting, the pressure will increase on regulation, the pressure will increase that we have more live data, the pressure will increase on compliance.

So anyway, with data like slow closes and, and, and even understanding what's in your bank account or not even the bank account, but your real like cash balance, the combo, your bank and your GL transactions and looking at what you really have access to is for capital increase.

So we'll need to use the technology to get rid of the task that.

You do capital and, and I think this causes a lot of inefficiencies in business.

8 hours a day where you don't really use enough your brain. And if you get rid of these tasks or you do less of them, then you have more time to learn, more time to think. And I think people will have basically more added value, will have more expertise. And I think we should ambition as finance to go towards that to also if you are a manager, do expect that your team is learning constantly. Do expect that as a part of their job your team should reduce the main task and augment the tasks that add value. And it should be like not a project of three months. It should be always like a continuous improvement. And if we don't want to do that in finance, then other people will come to us to put the pressure. So it's better that we drive it. And I'm sure we have, we're in the best position to do that because we have a lot of talented people. And I mean you can think of all of the profile. We have a lot of people. Often they also go to other departments because they are really good employees. They know a lot the company, they know the tools. They are really good also at communicating. So we should be proud of finance and have the ambition to push again the value we are bringing Business in general.

And if we can make companies 2% more efficient or 3% more efficient.

I love that, I love that vision. Okay, I know we're approaching the top of the hour.

We've got some great compounding impact to that. And if we can get data in front of the executives a little more quickly, I think that has a major, major impact on how businesses are run. Historically, the best businesses, the biggest businesses have been laser focused.

Questions in the chat. Are you okay if we run over a couple of minutes so we can get to some of those, those awesome.

I'm good. And, and I love super Accountant. Like the age of the super accountant. I love. Yes, it's happening.

It's what we are, we're all looking forward to, I think. Okay, Nicholas, China asks about Python and she asks if it's a program or an app and can you elaborate on how it benefits small businesses? And then Adil also asks how easy is it to learn Python for non technical people?

Yeah, so actually I have in two hours my course where we teach how to use AI and Python. But even if you never learn coding and I am the first example that you don't need to learn how to code, One year ago I got the bad chance to receive a letter from the tax department of Germany asking an audit on my sales because I was having too much sales outside of Germany and where I don't have a lot of sales tax on it.

So they were really on their accounting and I think delivering that enterprise quality of accounting.

Curious about where do I pay my sales tax and why I don't pay anything in Germany. And so then I called my, my tax advisor who is also doing the accounting and he tells me that now because of this we need to take all of the history of transactions and instead of booking one amount every month, we are going to have to book.

All of the trouble to small businesses and medium businesses will have a major, major impact on us over the long term. I think. And I think I'm talking more the five to ten year range. I think now in five years I think we've got, there's, there's a lot, we're dealing with a lot.

Back to the accounting system. What doesn't mean is that we had 80 CSV files and each of them 100 lines at least to 500 lines. And the problem is this CSV file we cannot upload like this in the accounting system because accounting system only recognize one column absolute value. So if you had negative value it could not work. So we had first to change that to make the column positive and have another column with the signed debit credit. And then we had to add account number, we had to delete some lines. So I was already like sweating and thinking it's going to cost me like one week of work of my accountant, $100 per hour. So then I thought, I have AI here. I'm actually using AI in Python also in my course I'm teaching that. So let's see if I can combine these 80 files in one. I went into AI, into HGPT and ask and it did that in my own environment, took the code, applied it.

And we're moving 100 miles an hour to try to figure out the absolute best ways to leverage AI.

But I the 80 files just became one. It was just in two minutes. So then I was, okay, now I could go in Excel and do all of this cleaning of data. But I thought that's going to take me half a day, let's continue and see if I am more ambitious. And I will go and try if I can continue to clean the data thanks to Python. And then I ask, okay, I think.

It'S that 5 to 10 year range where we're going to see some compounding impacts of this new technology. So again I'm very, very excited. I'm very optimistic about the future here. So sorry, with that I'll turn it over to you. I'm excited to hear your take on that.

And you change this column and make it positive and if it's negative, show credit. If it's positive, show debit. Again, it worked. Then I say, can you delete the PayPal lines? It worked. After 20 minutes I called my accountant, sent him this new CSV file he uploaded live with me and said everything is booked, everything is green. And the good news is we passed the audit without problem. So thanks to ChatGPT and Python and to show to everybody, how does it look like? Because you might wondering, I did a mini example less complex where imagine you have two have a sales file you want to combine. Well here those sales are in my Google Drive and all of the sales file look the same. So just like typical sales file, one line per day. And so I have 12 sales files like this. So if I want to have one annual file, that's my dream really and I think, and I hope, and that's my goal is that we are gearing towards having super accountant where we are not compromised anymore. Where oh this to go in each of them, copy and paste into one file. Or I can do what I did for my tax problem, I go to ChatGPT and I ask can you combine this file? And the thing is, I don't want to give my data inside ChatGPT. So I just ask the code. And now that I have the code, this person does only accounts receivable. This person does only accounts payable. I go to one environment that is called Google Colab. It's basically just Google Doc but for coding. And you can see like it's not really scary. It's just a white page where I connect my Google Drive and you can see that when I open it I can find the cells file here. So my 12 cells file, now that I have them, I take the code. So I will copy the code, I will go back to Google Collab, I will paste the code, I will just change in the code to see, to say where is my data? And then I will press the play button. It connect person does only data entry. Like we should not have that anymore. And again to Google Drive. And now in the bottom you can see in 15 seconds it was done. And now I have my new file called combine file. And I never learned how to code and I did that without knowing how to code. I didn't put my data inside. People don't want that anyway. If I take one example, I had a team where we had 10 people, three people went to retirement within one month and I just took over the team and we had to re challenge the way we were working. Instead I just asked how to do it and now you can think about how to do that with more advanced topics and that's how AI can use, can help you, how Python can help you as well.

I love this example. So Nicholas, I, the other day in the morning I was up early and I we're working on, you know, this, this new product and so I decided to dig in and see if I could use AI to generate code. Now like, I think I'm good at like architecting solutions but we have some incredible developers that work for us and then they pick it up from there and then they, they, they've been, they've developed everything, you know, and this is how it's been for six years. And I have never written a line of code in my life. And so I think the nice thing about AI is I would ask a question, it would give me an answer that I didn't understand and I'm like, okay, explain it to me like I'm a 4 year old and continue down that path and it will guide you down the right path to Getting the solution done.

And so of having people having just a focus on accounts payable. Then one month after, she was also helping on the closing and for her.

I showed like, I wrote code and deployed it and it worked. And I showed like at the beginning.

There was a bit of doubt, a bit of like, oh, like how am I going to deal that I only did the last 15 years accounts payable. But if you take the time and.

If our president or my co founder and I showed him this code and he like, it blew his mind that I was able to write. It was the first line of code I'd ever written. And I think if he'd read through my, my chat history, he would not have been as impressed. But like, it doesn't matter. Like this is where you know, hallucinations or data or whatever. Like this is one area where you can dive in and with zero knowledge, like deploy code and write code and solve problems like this. So I love, I love your example. You're obviously a lot smarter than me, Nicholas, but like even I was able to deploy code. So yeah, it was awesome.

That is to show how to also use technology to have more time during the day to do this additional task.

So cool. Okay, a couple more questions. We have a lot. I don't think we'll get to them all, but I'll try and get through as many as we can. So Selena asks, does NET Gain have any plans to involve AI with Net lease for lease accounting?

Then this person like just was so happy to do something new to.

Yeah, we do. Yeah, we've got a full. So we've got the full lease abstraction tool.

Also understand better why she was doing the task from before because she could see the end by seeing the closing. And I think we, we should really go towards that. Go to the fact that in that.

Lease that will read the lease agreement and even populate the fields into the lease record. Of course. Yeah, there's just, just the review and owning, you know, the accounting behind the lease before, before you finally hit submit and go. So it removes that, that lease abstraction component that has been so time consuming for, you know, especially during the adoption, but even now when you get a new lease. So yes.

Awesome. Thank you. Okay. Connie asks, what advice would you give to young people who want to study accounting or finance knowing they will enter the job market in five or 10 years.

So what you need to understand is first the three financial statements like, you cannot. We have super accountant that outside of accounting enter the job market without knowing that. And I've seen A lot of people when I was an auditor that didn't understand what is a P and L, a balance sheet, a cash flow statement and also how they connect to each other. If the pressure will increase on regulation, the pressure will increase that we have more live data. You know that normally you are ready to to work because you'll be already in the top 1%, the pressure will increase on compliance. So anyway it will increase. So we'll need to use the technology to get rid of the task that you do 10 or 5% compared to everybody else. Then everything we showed with technology, you need to build that on the side for sure. But that is not something specific to accounting. Every professional will have to build this technological aspect. But really I think accounting is a language that is logic and that you need to understand and you need to. It's not only learn, it's really understand. And once like you're in your mind like you have like the T account and that you have like all of the questions like coming, then it's good you are set to go. And this is like you have also accounting principles like they are. I publish a lot of about them on LinkedIn. Like there are 13 accounting principles. So really like learn that but make sure you understand it. Because if you are able to take any company and understand where is the cash impact, where there's not cash impact, where eight hours a day where you don't really use enough your brain. And if there is maybe like problem of inconsistency between all of the three statements, you get rid of this task or you do less of them, then you have more time to learn, more time to think. And I think people will have basically more demands. I really think you will be on edge and you can work everywhere in finance then.

Thank you. Anything to add, Adam?

I agree. It's the basics, right? Like making sure you really, really understand the basics and the core behind value.

Will have more expertise. And I think we should to see.

That'S how we can sense like there may have been a decrease in cash, but it may have been a really good quarter or cash may have gone up, may have been a bad quarter. Right. Like understanding the interplay between the financial statements and how it plays out I think is like the core AI is.

Not going to replace that because AI might know that. But if you are not able to judge if it's correct, then you should not work in finance if you don't know the three financial statements. That's like it should be your first knowledge to have.

Yeah. And I also Think there's these terms like I've heard there's a scalpel accountant and there's a chainsaw accountant and ambition.

As finance to go towards that to also material.

Materiality like really does matter in our world. And so understanding the big picture without a lot of us. Accountants are very detail oriented and we want everything to balance to the penny. But for a company we need.

If you are a manager, do expect that your team is learning constantly. Do expect that as a part of their job your team should reduce the moment task and augment the tasks that add value. And it should be like not a project of three months. It should be always speed and we.

Need, we need it to be materially accurate. And I think understanding that becoming a chainsaw accountant and like really producing a lot at a very fast rate I think is key. And technology is the big enabler. So obvious. The obvious answer is like go deep on the tech side in your preparation, become an accountant.

Awesome.

Thank you both. I know that we are getting short on time here so what we will do is wrap this things up in just a moment here. Any other closing comments that you two would like to mention as we finalize.

From my side? I think it's a very exciting time to be alive to be an accountant. These, some of these things that we've talked about today. It was great to see some of Nicholas's like very, very practical examples. I know were one of many, many software providers that are like going headfirst into the AI space and delivering up.

These applications like a continuous improvement. And if we don't want to do.

That in finance then it's an exciting time. I think there's a lot of noise. I think there are a lot of people talking about AI that are not specific and they're not talking about the practical applications. So I would be skeptical of the buzzwords but, but very, very go very, very deep on, on the practical applications here.

Other people will come to us to put the pressure so it's better that we drive it and I'm sure we have, we're in the best position to do that because we have a lot of talented people.

And again it's an exciting time and it's moving fast. Like two weeks from now, well, this will be obsolete. Right. We will have new tech coming in at us.

Right.

It's coming so fast it's easy to get left behind. So yeah, exciting time.

So for everybody that joined, I mean you can think of all of the profile. We have a lot of people often they also you are Already like in the top 1%, because you understand how important it is and you have to think a lot of people don't understand that or don't take the time to learn it or to discuss about it. So I think there you should already consider yourself in the top 1%. And also in finance, you have to think that we have access to a lot of data, we have access to all of the departments, we have direct access to management, where we see the strategy, we also discuss with them and we can influence them. And really, we have been always really good with technology, with Excel, Even in Excel, VBA, with the ERPs, often we do the implementation now with Power BI. And that's why now is AI. I really think that makes us the best at being the AI champions of the business. So if you are here today, have this ambition for 2025, become the AI champion of your business and also make sure you bring everybody else, everybody who was not there today or go to other departments because they are really good employees. They know a lot the company, they know the tools. They are really good also at communicating. So we should be proud of finance and have the ambition to push again the value we are bringing. The people from your team that are still skeptical, spend a lot of time, educate them, bring the messages that we gave you today to them. Because we also need to make sure everybody comes with us to continue and use AI in finance to add more value as a finance function.

Thank you. Great insights. Thank you both. Thank you to our speakers, Adam and Nicholas. Such valuable insights that you both shared today. I'm sure everybody found it very useful. Thank you all for joining us today on this discussion. Loved hearing your questions and insights. And as a reminder, we'll be sending out a recording of the webinar with additional resources. So if you have any further questions, feel free to reach out to our team at Net Gain or to Nicholas. And we look forward to helping you explore the possibilities of AI and accounting. Thank you so much, everyone.

Excellent. Thank you.

I love that. I love that vision. Okay, I know we're approaching the top of the hour. We've got some great questions in the chat. Are you okay if we run over a couple of minutes so we can get to some of them?

Those.

Awesome.

I'm good. And. And I love Super Accountant. Like the age of the super accountant. I love. Yes.

It'S what we. We're all looking forward to, I think. Okay, Nicholas, China asks about Python and she asks if it's a program or an app. And can you elaborate on how it Benefits small businesses. And then Adil also asks how easy is it to learn Python for non techniques technical people?

Yeah, so actually I have in in two hours my course with where we teach how to use AI and Python. But even if you never learn coding, and I am the first example that you don't need to learn how to code. One year ago I got the bad chance to receive a letter from the tax department of Germany asking an audit on my sales because I was having too much sales outside of Germany and where I don't have a lot of sales tax on it. So they were really curious about where do I pay my sales tax and why I don't pay anything in Germany. And so then I called my, my tax advisor who is also doing the accounting and he tells me that now because of this we need to take all of the history of transactions and instead of booking one amount every month, we are going to have to book all of the transactions back to the accounting system. What doesn't mean is that we had 80 CSV files and each of them 100 lines, at least to 500 lines. And the problem is this CSV file we cannot upload like this in the accounting system because accounting system only recognizes one column, absolute value. So if you had negative value, it could not work. So we had first to change that to make the column positive and have another column with the signed debit credit. And then we had to add account number, we had to delete some lines. So I was already sweating and thinking, okay, it's going to cost me like one week of work of my accountant, $100 per hour. So then I thought, I have AI here. I'm actually using AI in Python. Also in my course I'm teaching that. So let's see if I can combine these 80 files in one. I went into AI, into ChatGPT and Ask and it did that in my own environment, took the code, applied it, the 80 files just became one. It was just in two minutes. So then I was, okay, now I could go in Excel and do all of this cleaning of data. But I thought that's going to take me half a day. Let's continue and see if I am more ambitious and I will go and try if I can continue to clean the data thanks to Python. And then I asked, okay, can you change this column and make it positive and if it's negative, show credit. If it's positive, show debit Again, it worked. Then I say, can you delete the PayPal lines? It worked. After 20 minutes I called my accountant sent him this new CSV file he uploaded live with me and said everything is booked, everything is green. And the good news is we passed the audit without problem. So thanks to ChatGPT and, and Python and to show to everybody, how does it look like, because you might wondering, I did a mini example less complex where imagine you have to have a sales file you want to combine. Well, here those sales are in my Google Drive and all of the sales files look the same. So just like typical sales file, one line per day. And so I have 12 sales files like this. So if I want to have one annual file, I need to go in each of them, copy and paste into one file. Or I can do what I did for my tax problem. I go to ChatGPT and I ask, can you combine this file? And the thing is, I don't want to give my data inside ChatGPT, so I just ask the code. And now that I have the code, I go to one environment that is called Google Colab. It's basically just Google Doc, but for coding. And you can see like it's not really scary. It's just a white page where I connect my Google Drive and you can see that when I open it, I can find the sales file here. So my 12 sales file, now that I have them, I take the code, I will copy the code, I will go back to Google Collab, I will paste the code, I will just change in the code to say where is my data? And then I will press the play button, it connects again to Google Drive. And now in the bottom you can see in 15 seconds it was done. And now I have my new file called Combine file. And I never learn how to code and I did that without, without knowing how to code. I didn't put my data inside ChatGPT, I just asked how to do it. And now you can think about how to do that with more advanced topics. And that's how AI can use, can help you, how Python can help you as well.

I love this example. So, Nicholas, I, the other day in the morning I was up early and I, we're working on, you know, this, this new product and so I decided to dig in and see if I could use AI to generate code. Now, like, I think I'm good at like architecting solutions, but we have some incredible developers that work for us and then they pick it up from there and then they, they, they've been, they've developed everything, you know, and this is how it's been for six years. And I have never written a line of code in my life. And so I think the nice thing about AI is I would ask a question, it would give me an answer that I didn't understand. And I'm like, okay, explain it to me like I'm a 4 year old and continue down that path and it will guide you down the right path to getting the solution done. And so I showed, like, I wrote code and deployed it and it worked. And I showed our president, my co founder and I showed him this code and he like, it blew his mind that I was able to write. It was the first line of code I'd ever written. And I think if he'd read through my, my chat history, he would not have been as impressed. But like, it doesn't matter. Like this is where you know, hallucinations or data or whatever, like this is one area where you can dive in and with zero knowledge, like deploy code and write code and solve problems like this. So I love, I love your example. You're obviously a lot smarter than me, Nicholas, but like, even I was able to deploy code. So yeah, it was awesome.

That is so cool. Okay, a couple more questions. We have a lot. I don't think we'll get to them all, but I'll try and get through as many as we can. So Selena asks, does NET Gain have any plans to involve AI with Net lease for lease accounting?

Yeah, we do. Yeah, we've got a full. So we've got the full lease abstraction tool in that lease that will read the lease agreement and even populate the fields into the lease record. Of course. Yeah, there's just, just the review and owning, you know, the accounting behind the lease before you finally hit submit and go. So it removes that, that lease abstraction component that has been so time consuming for, you know, especially during the adoption, but even now when you get a new lease. So yes.

Awesome. Thank you. Okay. Connie asks, what advice would you give to young people who want to study accounting or finance knowing they will enter the job market in five or 10 years?

So what you need to understand is first the three financial statements. You cannot enter the job market without knowing that. And I've seen a lot of people when I was an auditor that didn't understand what is a P and L, a balance sheet, a cash flow statement, and also how they connect to each other. If you know that normally you are ready to work because you'll be already in the top 1% or 5% compared to everybody else, then everything we showed with technology, you need to build that on the side for sure. But that is not something specific to accounting. Every professional will have to build this technological aspect. But really I think accounting is a language that is logic and that you need to understand and you need to. It's not only learn, it's really understand. And once like you're in your mind like you have like the T account and that you have like all of the questions like coming, then it's good, you are set to go. And this is like you have also accounting principles like they are. I publish a lot of about them on LinkedIn. Like there are 13 accounting principles. So really like learn that but make sure you understand it. Because if you are able to take any company and understand where is a cash impact, where there's not cash impact, where there is maybe like problem of inconsistency between all of the three statements, I really think you will be on edge and you can work everywhere in finance then.

Thank you. Anything to add, Adam?

I agree. It's the basics, right? Like making sure you really, really understand the basics and the core behind it. To see that's how we can sense there may have been a decrease in cash, but it may have been a really good quarter or cash may have gone up, may have been a bad quarter. Right. Understanding the interplay between the financial statements and how it plays out I think is the core.

AI is not going to replace that because AI might know that. But if you are not able to judge if it's correct, then you should not work in finance. If you don't know the three financial statements. That's like it should be your first knowledge to. To have.

Yeah. And I also think there's these terms like I've heard there's a scalpel accountant and there's a chainsaw accountant and material, materiality like really does matter in our world. And so understanding the big picture without a lot of us accountants are very detail oriented and we want everything to balance to the penny. But for a company we need speed and we need it to be materially accurate. And I think understanding that becoming a chainsaw accountant and like really producing a lot at a very fast rate I think is key. And technology is the big enabler. So obvious. The obvious answer is like go deep on the tech side in your preparation to become an accountant.

Awesome.

Thank you both. I know that we are getting short on time here, so what we will do is wrap things up in just a moment here. Any other closing comments that you two would like to mention as we finalize.

From my side? I think it's a very exciting time to be alive, to be an accountant. These, some of these things that we've talked about today, it was great to see some of. Nicholas is like very, very practiced examples. I know we're one of many, many software providers that are like going headfirst into the AI space and delivering up these applications. It's an exciting time. I think there's a lot of noise. I think there are a lot of people talking about AI that are not specific and they're not talking about the practical applications. So I would be skeptical of the buzzwords, but, but very, very go very, very deep on, on the practical applications here. Again, it's an exciting time and it's moving fast. Like two weeks from now, this will be obsolete, right? We will have a new tech coming in at us, right? It's coming so fast, it's easy to get left behind. So, yeah, exciting time.

So for everybody that joined, you are already like in the top 1% because you understand how important it is and you have to think a lot of people don't understand that or don't take the time to learn it or to discuss about it. So I think there you should already consider yourself in the top 1%. And also in finance, you have to think that we have access to a lot of data, we have access to all of the departments, we have direct access to management where we see the strategy. We also discuss with them and we can influence them. And really, we have been always really good with technology with Excel, with, Even in Excel, VBA, with the ERPs. Often we do the implementation now with Power BI, and that's why now with AI, I really think that makes us the best at being the AI champions of the business. So if you are here today, have this ambition for 2025, become the AI champion of your business and also make sure you bring everybody else, everybody who was not there today, or the people from your team that are still skeptical. Spend a lot of time, educate them, bring the messages that we gave you today to them. Because we also need to make sure everybody comes with us to continue and use AI in finance to make add more value as a finance function.

Thank you. Great insights. Thank you both. Thank you to our speakers, Adam and Nicholas. Such valuable insights that you both shared today. I'm sure everybody found it very useful. Thank you all for joining us today on this discussion. Loved hearing your questions and insights. And as a reminder, we'll be sending out a recording of the webinar with additional resources. So if you have any further questions, feel free to reach out to our team at Net Gain or to Nicholas. And we look forward to helping you explore the possibilities of AI in accounting. Thank you so much, everyone.

Excellent. Thank you.