8 ways that MindBridge Ai Auditor is revolutionizing financial auditing
Well, good morning, good afternoon, and good evening to all of you around the world. Getting ready for another exciting webinar series from MindBridge. This time talking about the "8 Effective ways that the MindBridge Ai Auditor is Revolutionizing Financial Audits." I'm JC, and I'll be your host and guide throughout this tour, for the next little bit of the webinar. So let's get going. Financial audits. It's a piece of review work that happens every year for most organizations, sometimes multiple times throughout the year.
Sometimes, people wonder why it happens, but at the end of the day, it happens. All organizations have to actually go through this process in order to help their stakeholders understand exactly what's going on within the financial context of the business. Now, at the summary level, people look at profit and loss statements, income statements, balance sheets, statements of free cash flow, and they make various decisions based on those statements that they see.
But it's what's underneath the covers that really concerns a lot of auditors and a lot of accountants, to make sure that they're accurate. And, as you can see here, for instance based on the citation from ICAEW, it's about truth and fairness and that they were properly prepared. So that begs us to the question of how these things get done and what are the issues that are currently plaguing the system.
Well, it really comes down to a couple of challenges that had been in the audit space for a number of years. The amount of data has continued to increase year over year, as bigger and bigger organizations get started, as more omnichannel approaches come to business, where they're not just selling, you know, direct to consumer or business to business anymore, they might be doing both.
And they might have huge distribution networks around the world trying to carry their offering out to the market. So the ability to really understand that financial health of an organization, based on the data, has really become quite challenging. We still have some decades old techniques also in use. And while we're very, very excited that there's people continuing to flood into the CPA profession, the reality is taking the time to build the subjective and judgmental based awareness over the years takes a long time.
And if all you're doing is sampling, then you potentially are going to miss something that's really material. Now, part of that is that the limited human capacity that we talk about every once in a while, which is when you think about how often mistakes can happen, and trying to find every single one of them, that needle in the haystack of all the tens of thousands or hundreds of thousands of transactions an organization might do on a month or quarterly basis, you would have to essentially deploy the same number of accountants and auditors to audit the books, as is making the entries themselves.
And that time and effort that it would take is likely improbable to have anyone to pay for. So what happens is we get this process that, although has been very well used over the last few decades, it's actually not doing as good of a job as it can to find those errors or find those issues within the financial statements.
And it's so drastic that, in fact, in the last 2 years alone, in the Report to Nations published by the Association of Certified Fraud Examiners, that the estimate has increased from in the low $3 trillion mark to $4 trillion or more, rather, of the likelihood we're not finding different types of financial loss.
Now, financial loss, when it's due to human error, is one thing. When it's an intentional thing, then we're talking about fraud. And that $4 trillion, if you start putting it into context for your pocketbook or for investors of the organization's pocketbook, that could be as much as $550 or so for every person on the planet if they had a stock portfolio, just being evaporated throughout the system.
So MindBridge's mission is really to help auditors, accountants, and teams find any and all issues that it possibly can inside the data actually help present what we think are the most likely culprits in an organization of having some type of issue, hopefully all of them human error and not by something intentional.
When we think about just how much of an issue this presents, we have to talk a little bit about the tools and techniques of finding these things. And like I said at the beginning, it comes down to we just have to look at all the pieces of the puzzle. So fraud analytics catches up to 3% of that $200 billion known knowns.
The vast majority and the highest-ranking contributor to finding the fraud for some of these errors is simply by a tip line. Think of the old facsimile lines, where people are just sending in a fax saying, "Oh, I think someone's done something wrong."
That's what we're talking about. And the time it takes to detect the fraud usually spans multiple quarters, if not multiple years. And when we think about how we try to solve this problem, it really is the fundamental position of reasonable assurance, is built on doing sampling. And sampling is really just a coping mechanism for us, human beings, to say, "We can't reasonably be able to be asked to look at every single transaction."
Because I mentioned, if you wanted to do that, you would need the same amount of accountants and auditors and bookkeepers all looking at the transactions that have already transpired. And that's just too heavy for an organization to do. So I guess, at the end of the day, what I'm saying is the current methodologies need an update and the current tools equally need an update to help solve this problem.
And over here at MindBridge, we spend a lot of time thinking about how we can solve this for you and how we can help participate in the answer. And it really starts from smartly ingesting the data from anywhere, from any subsystem, and immediately giving insights and values to you, the users, so that you can actually dig in as deeply as it makes sense, giving your engagement or your level of work that you're planning to do with your clients.
And from there, make sure that you can actually track it and have the reports that are required to get back to that point of reasonable assurance and having gone through all these steps, to make sure the financial statements are in fact, you know, truthfully and fairly represented, and that they've been prepared in a good sense.
So we spend all of our time thinking about how to help you. And we've tried to narrow it down to just eight simple ways that we think we're revolutionizing the financial auditing experience. And really, what we look at is, first and foremost, the user experience. The last time I looked at the books, the various courseware and books that are in most accounting curriculum at various universities.
We spend a lot of time learning tools like Excel, learning various manipulation methods. And we have a fundamental belief that, if you have to learn how to create pivot tables and scripts and, you know, work through the data just to create the groupings and make sure that everything is right, you're spending an awful lot of human capacity on something that machines have been doing for multiple decades now.
Now, they may not have known accounting, but maybe today they do. The other thing that we think about a lot, and one of the ways that we think we're really going to be radically changing and helping shift the marketplace, is the fact that sampling is still going to be done, but could it be done based on having an unbiased system reviewing 100%of the data before you pick that sample?
And this is something that we really feel very strongly about. Every bit of data that is maybe important to the business should have a chance of being reviewed or selected. I mean, that's right in all of the international standards for auditing. And, in every single geography, the standards both from a public company oversight, as others, wants to make sure that every transaction has that chance.
But what if you're able to actually identify the ones that are potentially more material or have a higher degree of confidence that there's something wrong with them? Even before that, getting 100% and then analyze and then picking the sample. That might be better. Now, I'm going to switch a little bit between, you know, thinking about it more from the high-level pictorial view, and I'll actually dig right into how the Ai Auditor can help.
So let's just take those first two points. The ability to grab data and analyze it and assess it, with zero scripting, and getting 100% coverage. So how can we do that? Well, let's start with, let's say a basic bit of auditing, which is to look at the general ledger details. Part of this is that, obviously, you have to go through and look at all of the data as best as possible.
You can grab in the opening balance and closing balance, but let's just work on the premise that there's just simply a set of data we need to go and graph. Now, we do connect directly to some of the leading cloud ERPs, literally log in, hit go, select your time frame, and it will execute all of the analysis. But more often than not, if you're in a CPA firm, you're a public accountant, you're a consultant, you're typically getting some type of file from their ERP.
And sometimes, they're quite messy. But what you get to do with the Ai Auditor is you actually get to just drag and drop various data elements in. So what do I mean by that? Well, let's just take a data set and let's drag it in. The data set I'm representing here in the background, the Excel, just a simple transaction by account report showing me all the transactions for every single account.
Now, I'm sure most CPA firms have a lot of smart people, and they'd be able to just take that data, whip it around in some pivot tables, clean it up, make it really super simple, and that might take them a couple of hours to do. But what if they could do it differently? Well, with what we do, we give you that ability to actually go and drag that data in to the system.
So let's go ahead and do that. And let's pull that data out to the forefront. Just take a second for the system to show. I think the system's finally updating your view. So it's taken that data in. Now, what did it do with that data?
Well, it started dynamically building your account groupings. It started to dynamically look at the information and see what it needed to know about it. The first thing it tries to do is map all of those columns of data into the system. So it's telling me, "Oh, we knew what this was. It was group format that we're used to,"in this case Sage 50. And there's nothing else for you to do. Awesome.
Well, great news. But then it does ask me a few more questions. "Hey, in this engagement, are you looking at a particular materiality threshold? Let's throw 5,000 in there." Are there specific suspicious keywords that might be of interest for you to go look at, maybe based on your dialogue with the customer before you start it? Maybe we need to go and look at a particular customer name or a particular scenario that happened.
You know, maybe we're working on a big relocation, and we want to make sure that nothing in that relocation has caused some type of error, because it's a really big ticket for that company. All we have to do is click on those few pieces, and automatically, the system is going to go and do its work. Now, it's going to tell me that the accounts that haven't been verified yet, it's going to ask me where they go. So in the time it took me to grab that data in and get to this screen, it's already built me a dynamic account hierarchy.
Now, of course, you may want to import an account grouping or an account mapping from another tool, if you're using an engagement tool that has a working paper solution where you've already done this work, no problem. Just use the import/export button on the side. But if this is the first time it's seen the data and it's never seen anything else from you, it's simply going to do the work and all you have to do is verify.
Once verified, the very next time you go through the system, the very next time you load data, maybe in the next calendar year when you're doing their next audit or their next interim period, it will just show you the ones that haven't been verified that are new from one year to the next. And then simply put, once we have those pieces in place, all we have to do is start the analysis. Now, I'm not going to load in the closing balance and the opening balance.
I'm just going to hit go, because I want to show you just how significant and how quick this work happens. Now, this will work in the background as I bring up my next two of the eight ways in how we're revolutionizing the audit. So we talked about the 100% data analysis. We talked about how quick and simple it is with zero scripting needed.
So what the system is doing, as we're talking, is it's going through and it's going to assess every single transaction, every single entry within the transaction, and it's actually going to score it in the confidence level between and 100 of whether we think there's a risk for an error, from an unbiased machine view.
Now, what that means is take your weekend posts, take your manual entries, your top N%, 2% material values, my material threshold of 5,000, and let's start getting a little bit more sophisticated. Let's do some regression testing in there. Let's start doing some Benford's law. And oh, by the way, there's some AI baked in here too, which I'll talk about more in a second. And it's running every single one of those unique tests against every single entry and transaction, and it's cross-correlating that across every other transaction in the data.
And it's using a known known or an exemplar database to actually validate against the patterns that we've known and seen to be either an error or a fraudulent transaction in the past. And it's doing that in minutes, and you'll see this as we move over. So taking all of your data, bringing it in, and being able to see results.
And those results, really, at the end of the day, are two-fold, and we'll talk about the predominant one that we like to focus a little bit more on, which is identifying the risk. If we go back to the ICAEW statement about it being about preparation, and truth and fairness, and, you know, making sure that we follow along the lines of reasonable assurance as well, what we're trying to assess is where the potential risk is that an auditor or an accountant needs to spend the time.
And we're going to do that in seconds. And we're going to be able to show them not just the high level view, but we'll show you throughout the rest of this presentation how we go deeper and deeper and deeper to get new information. So the results are completed, as you can tell, in that time it took me to run two quick slides, and it's giving me a first pass of picture as to where some of that risk assessment is.
So I can very quickly see, without even having to dream up what it could be, I can see where I need to focus my time. I've got 7 transactions here, roughly $3 million in monetary flow, that are flagged as high risk, which means those control points that I've triggered, be it that it was a weekend post, compounded with a suspicious keyword, compounded with the fact that it was a reversal, compounded with the fact that it was a failed statistical model like Benford's law, by our expert score, and I can go on and on all through our control points.
It was seven transactions that had a high degree of confidence that they're worth looking at. It doesn't mean that there's an error, but there's highly likely a reason why you'd probably want to include it in your audit. Then there's 56 more that are in the medium bucket, almost $12 million of flow, and the balance is sitting in low. Now, as I said, that's just a high level picture.
How do we get it a little bit deeper? Well, right underneath that, we can actually look at things by month, day, or week, depending on the granularity of the data that you give us and the periodic view you want to have. You can go down to any level of detail. I can remove all the green noise of what's considered and classified as the low risk. And I can just look over time, slide this rule, and find, "Well, where and when are these issues happening?"
So right here, week 50 to week 52 and week 1, so a crossover in a calendar year, which is also a fiscal period for this particular organization. I may want to spend some time there. But just knowing when you need to review and what sort of possibility you might need to look at for cutoff periods, etc., we also give you this amazing risk burst that kind of radiates out from the core account classifications out to some of the level four details where we think risk is.
So as I slide into the first slice here, I can see I'm in assets, current assets, accounts receivable. It's got this really nice rich green. This rich green is telling me that, you know, relatively low risk. But even within that assets classification in my current assets, I've actually got a bit of a problem here.
Under my other tangibles, I can see that I've got equipment as about $730,000 worth of monetary flow that the system is saying, "You know, there might be a little bit of risk there." But I'm sure, when I first scrolled up to let you see the risk first, everyone's eyes through right to the oranges and reds. Oranges and reds, just like a traffic light, you better stop, take a look.
And what this is telling me is that the director's fees, management salaries, and bad debts, roughly about $1.5 million of flow. If I added those three pieces together, look like they might be having some issues. We might want to spend time there. So, again, our goal as we look at it, what we're trying to do is help inform.
You know, a big challenge in the financial audits today is that, based on the fact that we cannot conceivably or reasonably go through every single transaction manually, by human eyeball, we take a sample. We're giving you guidance on where we think that sample needs to be. Now, that's not going to override yourselves and your professional judgment.
That's not supposed to be override any other procedural guidance you're getting from whatever standards body you follow around the world. It's more about partnering with you. It's more about bringing a different level of focus and allowing you to continue to assess as you're going through. So this is what we do. This is how quick we got to the information. This is how quickly we were able to assess the risk.
And we were able to get going a little bit further. Now, there's a lot more to it than just identifying risks. There's also the fact that we want to show you exactly and explicitly where these errors and anomalies are. And I've started to tease you with the fact that you can look at it by time, you can look at it by type of accounts, specific part of the grouping and mapping.
But the reality is you want to be able to interact with this and find out exactly where those issues are. And without any scripting, without any trying to, you know, write your own codification, you can actually just ask the system, "Hey, what things triggered Benford's law? What was also in that population that's manually entered? Did it fail the expert score?" And so on and so forth. And you'll get to a really great list.
And underneath that, down at the weed level, down at the entry level, well, actually, it can show you exactly which ones have a challenge. And it's all about starting to bring new insights into the fold as well, because a big part of bringing audit data analytics to bear and to be really well-defined in your approach at helping your clients exceed their expectations and increase value, it's about also sharing with them insights as to where they need to be focused.
You know, it's not just to do the tick box and say that the financial statements are, you know, clear or there, or you're reasonably sure they are. You also build that audit opinion about the key factors that you sought through. Well, why not pair that with the manual work that you're going to do anyways, combined with some of these new insights.
So what are some of these insights that we can bring to bear? Well, if we go back to the solution just for a second and we look at all the information that's available, it really can give you that many more insightful things. So first and foremost, in that same time that it took us to load the data in, to do the account grouping, to then do the analysis, the systems also pre-prepared the core financials, balance sheet and income statement, and it's also gone through and built a whole bunch of key performance indicators or key ratio reports for you, based on the data that it has.
Obviously, if you're working on an engagement, you'll probably want to add in accounts payable and maybe some other data. And this list of reports will be a little bit bigger, and it will have more things. But the point is that it dynamically creates these, so that what you can do from that point forward is just validate against what your client's already given you, at a certain level of granularity. How great is that? But sometimes, just looking at Excel reports gets a little bit dry and boring, so you might actually want to go to a more interactive approach, where you can actually visualize how things are trending over time.
And not just how they're trending over time, because I think there's many visual analytics tools that can do that, but here's where we pair it with that risk perspective, which we think is supremely different and very revolutionary in the way that we're thinking about, you know, the world and the future, as we think about the audit of the future.
It's about looking at all the factors that potentially influence the goods, the bads, the uglies of those financial statements. So we can look at the balance sheet, we can look at various classifications of things, we can look at maybe our payables and find out where there might be some risk over time. As you can see, the data set that I loaded in has three years' worth of data, or two and a half years.
We haven't quite gone through the full year yet. So obviously, it's tailing off there, but you can see the peaks and valleys as it relates to the payables. And underneath, you see the risk comparison right away. Again, let's remove those lows and let's focus on, period over period, where are the potential risk points? How many entries or transactions are at risk? And then if I scroll a little further, you'll see a few more things that are obviously very important, days expenses in AP, how is that trending, different things like days sales outstanding, and again, various elements of things, like inventory and others.
It's about taking all of this data very quickly and giving you something new and insightful that you can use with your client. And as I said, it's about finding some of those errors and anomalies that the human eye is not going to catch just by doing a scan, because there's just far too much data to know exactly where to hunt that for these types of things.
So what we do in our solution is we spend all the time working for you by having the solution do all of this with a high degree of confidence and showing you those confidence intervals of high, medium, and low, and getting even deeper. And I'll just touch on it a little bit quickly, and then we'll kind of come back to a few more revolutionary ways. So when I look at the data and I look at all of the information underneath, I can see very quickly the scoring ranked high to low of all 2,966 transactions, or if I flip this to entry mode, I'd see 13,000 and some number of entries.
And not only do I see, you know, when they were posted, I see the monetary value and how many entries are part of it. I can also dig in and see what's happening. So I can see here that this particular transaction appears to be the result of some type of accident, a fire, some type of incident number.
So this is good information for you to have right at your fingertips. You don't have to go and ask the client to bring you any more information. Some of it's just right here. But when we look for anomalies, when we look for things that might be erroneous, what we really want to do is actually then dig in and understand why. Because one of the things that's happening in the marketplace as people talk about this new thing of artificial intelligence, which has been around for almost 60 years but seems to be making quite the breakthrough again, is the fact that it needs to be explainable.
It cannot be a black box. We need to be able to deliver you insights very quickly on where and what we think are potentially issues within these transactions. So as I've been maybe two clicks into the transaction, I can see exactly what the system thinks is potentially problematic. "Hey, it's a cash to bad debt conversion.It applies to another journal entry that's out there. There's last three digits in zeroes or nines for two of the entries. It's obviously over my materiality threshold."
But where it starts to get interesting is things like the outlier anomaly and the rare flows. These are machine learning algorithms that actually create an even further degree of confidence to know why something was selected. So this one has been selected because it's the largest outlier in the population of data that you're analyzing right now. So imagine 500 million transactions and entries, trying to find that needle in the haystack, trying to find out what the outliers are.
It can be very difficult, especially as the companies get larger, as their transactions get larger, as they become more complex. So what you're able to do is not just look for the needle in the haystack, but as one client said to us, "You're just going to burn all the haystacks and find the needles." We can give you that power as an auditor, as an accounting executive, as a CPA, to actually dig into the details in a multitude of ways, but not just to give you a good feeling that something's potentially an error or not, but actually give you actionable insight, "Wow, this is the biggest outlier.
Oh, there are 600 other transactions that are outliers that are above 50%. That's not good. We should probably look at those," right? Maybe not all 672 all at once, but you've got a team for that, right? So what happens is you get the ability to look at this. You see, "Oh, yeah, definitely something I want someone to go follow up on." I can very quickly go ask one of my team members to go follow up on it.
We'll send it over to Sam, give him a due date of Monday, you know, why not make him work the weekends. And away that goes. Sam's now going to get notified. And greatest part about this solution being cloud-based, ready to go anywhere you are, as long as you've got a browser, whatever Sam is, Sam is getting a notification.
He can log right into the system. He can see exactly what is asked. So imagine all of those status calls that you do on a weekly basis or more frequently during your two, or three, or four-week request audit with your client. You can be communicating with your team, whether they're on site or off site, whether they're at different sites, because you're working on an engagement that has four different subsidiaries, all communicating in harmony.
Finding the same insights, looking at the same information, being able to follow it up in a certain way. So we give you quite a bit of different views into the data and help you discover those new insights and find errors and anomalies. And I think that in some ways, when you think about what that then drives to is this concept of redefining reasonable assurance.
And when we think about what's possible today, with the advent of machine learning algorithms, of artificial intelligence, which is really an umbrella that includes machine learning and other techniques, natural language processing, natural language understanding, and a few other things, I think what's reasonable yesterday is different than what's reasonable today. Because there's now tools, explicitly the first one out to market and I think the only one today, being the Ai Auditor, where it goes through 100% of your transactional data, 100% of all the information, cross-correlates it against each other, looks at things like monetary flows between the data sets, all with the click of a button, using a lot of these AI techniques.
And so, what's reasonable today is to think that you might just need to go a little bit deeper using a machine as your partner in this category. And as I've mentioned a couple of times, bringing in machine learning and AI. So you're starting to see a little bit of it. I'm going to go back into the software and show you a little bit more about how the tool sort of segregates data into things to be thinking about as you think about this new world of AI as part of your partner in crime, oh, that's probably a bad word, your partner in audit, is to obviously see how this can help you.
How can it aid you to get the best information possible, not only for your client but really to formulate the right audit opinion at the end of the day, if you're using this for a stringent financial audit? So as I said, go back in and let's explore what I mean by that, leveraging machine learning, leveraging this new concept of maybe having a different sense of what reasonable is.
Let's go through how we're going to get the job done. So this is all great and dandy. I've isolated one example, one transaction. But really, our jobs are much bigger than that, right. So we start from left to right, again. Where's my risk? I see my risk.
I now understand a little bit better based on the control points that it's triggering. This is how we classify risk. And you can almost think of it like a scale. Everything's on this weighted scale. If something is a manual entry, it may not be that bad. If something is a manual entry and it was done on the weekend, it's a little bit worse than just a manual entry, but it may still not be bad.
Maybe it happened at the beginning of the year. These are all rules that you're testing for anyways. But now, when you aggregate those up, you're like, "Oh, well, yeah, maybe that's a little bit more interesting." But now let's start to get to the really interesting things that come in when you start bringing machine learning and artificial intelligence. Just kind of highlight the one called the expert score.
The expert score here is flagged 128 transactions. Think of this like a having 1,000 other auditors side by side with you. The way that the Ai Auditor has been developed, it's got a number of different machine learning algorithms. One of which is a supervised set, which we call our expert score. And what that means is, when you hear the whole big term AI and people talk about machine learning and building neural net, etc., you're training the system.
Well, we've done that work for you. We've taken our half billion data points. We've taken a series of CPAs, whether they're chartered accountants from the UK, from Canada, from the U.S., whether they might be Certified Fraud Examiners, Certified Financial Crime Specialists. And we have essentially tried to emulate the way that they review transactions and give you a sense of how they would trigger them.
Are they interesting or not interesting? Are they likely to be an issue or not? And we've codified all of that for you over the last three years and bundled them into just one of our machine learning algorithms. The reason we have multiple is because we actually really feel strongly that you need to be looking at this almost from a 360-degree view.
You need a system that's going to be partnering with you to identify anything that's interesting about a single entry or a single transaction, to help you in getting that job done and helping you making sure that you are reasonably sure there aren't any errors or misstatements. So we use things like the expert score. We use things like outlier anomaly. We use things like the unusual amounts control point.
And when these trigger, these machine learning algorithms start triggering and there was a few more, they are a little bit higher on the weighting score. So when we say there are seven things that are important, you know what, they probably are most interesting, because it's using a wide collection of techniques, rules, stats, and machine learning, to give you a confidence score. So, again, how do I get my job done? I've done my planning and analysis.
I can see all the trends I need. I've got all my financial reports. At the end of the day, where we are right now with all the standards around the world is we still have to do substantive. We still have to pull sometimes that our team, whether they're juniors or seniors, are going to go and work on. And you want that to be as effective as possible. And even within things like the PCAOB Guidelines in the U.S., for any of you that are U.S.-based that might be listening in, there are clear and cut directional lines for you to say that you should enrich your data in the best way you can, and you should be filtering based on potential risk.
That's how the guidance tools are built, that's how all of these elements are built. We give you the next level of risk as part of this overall process. And what I mean by that is your guidance tool may say, based on everything you've put in on your judgment-based assessment of that business, "Hey, I need you to pick a sample of, I don't know, let's pick 30 from revenue." No problem. What's impacting revenue?
Boom, there's my list. Okay? Got 1,151. Let's go launch the sampler. It tells me that I need to pick 70 of them, let's say. It's not 700, pick 70. What the system is going to naturally do is it's going to take anything it can from the high risk, in this case 3, it's then going to take anything it can from the medium, up to 60% of the total remaining balance, and then the rest from the low.
And if you think about the guidance, you think about the guideline, you need to be able to say that everybody had a chance of being selected, absolutely done. You can use statistical sampling, absolutely done. You need to enrich the data with some level of risk assessment, boom, done, we're in. Generate that sample. That sample, now, can be assigned to one or many parties, and it has the ability to drive down to people working audit in process within your team.
But you know, sometimes, based on what you've learned, you actually want to go a little bit further than just saying, "I'm going to pick my 70 from revenue." Sometimes you actually want to be talking to the system and getting more insightful information as part of your sample selection, maybe because you're just unsure of something.
So maybe you want to show risky transactions that are impacting, I don't know, let's say cogs that get sold are over, let's say, $5,000 and experts don't like. I can ask a question like that of the system, and it will naturally come back with anything it finds within that spectrum.
And you'll notice that it's not just like your typical search engine. This is usually a combination of natural language processing concepts, and it's actually redefining your question or your statement into something that's fully clickable. So unlike a search engine, where you have to then retype it and say, "Oh, I really meant or," and then hit search again, it's just there.
It takes any of those conditional types of questions and allows you to just pick and choose. If I want to say, "Well, I really didn't mean that risky, you know, 50 to 100. I need a little bit more context, maybe 30, so medium and high, it's going to go and pull me that list as well.
So what it does is it allows me to grab a question or a statement or a thought that I have about this data, something I don't need to go and select on or sample on, and then you can still use our sampling to go and get those pieces of information. Let's say, in this case, it tells me I need 50. And now I can see that, "Oh, there's already two in my plan, so someone else is already working on those.
Great, let me take the remaining one." "Oh, there's 22 other transactions that are available, so obviously, let's go do that." And then it's going to take the remaining balance. So what you've done in a few short clicks is you've been able to interact with the system naturally. You've been able to refine what you're looking for based on the guidance tools or based on your own audit process and methodology.
And immediately, getting in there and working through the data itself, creating an audit plan, which then will be assigned to various people, and you'll be able to actually go and pick those pieces apart. So I can export this. I can send the notifications directly to the team. They can log in and away they go.
They can see the data. So now, when we say all of this...okay, let's recap for a second, eight ways that MindBridge Ai Auditor is revolutionizing financial auditing. I kind of feel like I've actually gotten past that. We've got everything from zero scripting to embedding machine learning to be able to easily identify errors or anomalies to finding new insights about the data by merging risk-based scoring combined with traditional KPI or ratio-based analysis.
We've dynamically ingested the data. We've created the account grouping without you having your hands on the data. I mean, I can go on and on on the list, and all of this is powered by machine learning and AI technology. Sounds pretty fantastic. What usually ends up happening when I get to this point, people have lots of questions.
So don't worry. Put them into the question and answer box, and I'll get to them. But if we think about what's happening in the industry, we think about what our leaders are saying. This statement I came across a little while ago, and I think it's really apropos to the conversation. We will see more change in the next 5 years than we've seen in the last 50, paraphrased a little bit there. Think about that.
We will see more change in the next 5 years than we will in the last 50. And this is a very important and critical piece of information, because from my perspective, the time is already here. That next five years has already started. The time started a while ago. And a big piece of our belief system here at MindBridge is based on a number of interactions with firms all around the world, but it's that we don't believe that AI is actually going to replace you or replace the CPA.
What's really important is that the accountant or the auditor and the CPAs embrace it. And those that do are likely going to replace those that don't. I think we should all be fairly real in thinking that the population coming out of school and staying in the profession is dwindling every year. And it's not just because of the tools or antiquated processes, but it's actually about trying to find a different balance.
You know, using newer tools, going a little bit deeper, making it about really providing client value is where it's at. And I think that we can effectively say that, you know, not that it's five years away, not that it's two years away, but that it's here now. There's a purpose-built platform which I've just walked in and out of as we've gone through this last 45 minutes or so, and all its job is to look at your payable subledger, general ledger, other subsidiary ledger data, looking at, you know, a vendor analysis, a customer analysis, employee analysis, and find anomalies, find issues.
It's using the latest technologies, but it's using the latest technologies purpose-built and defined for the industry. So we're not talking about, you know, the big conglomerates saying, "Hey, we've got this great technology platform. What do you want to point it to?"
We're saying, "Oh, we've trained it to help with audit." And so you're able to now get to 100% analysis on every single engagement. In minutes, it's done its first pass. In less than hours, your team can actually go through the rest and sample out and select out the things that you think need more work.
It's no longer a black hole of data prep time. It's no longer three, four, five days later, you've got financial statements to look at after you get the client's data. Load, see reports, start interacting, minutes. And what I really think of the last point, which is really key here, is the finance and accounting professionals have always been on this track, at least the last two decades that I've been spending time with them, to be more of a trusted advisor and more strategic advisor within an organization.
I think the same can be held for any professional in the accounting space, including the external audit teams. They're going to be higher and higher in the value chain if they can show more value at the end of the day. It's all about getting to a level of detail and a level of view that actually gives you more and more out of a system that is partnering with you and, dare I say it, creating audit cyborgs out of you, where you partner with an AI system, like Ai Auditor, to actually do a more effective and efficient job.
And maybe, just maybe, we'll move from the annual 15 months out of, you know, rear view mirror style of auditing, and we'll move into something more progressive, where it's continuous and it's, you know, quarterly or monthly. I think you'll be able to use technology like this to go even further and deeper, in forensic engagements, or looking just to make sure that in the internal audit process you're catching things more quickly that might be erroneous.
Because again, if you remember, it's the intent that makes it fraud, but there are still huge amounts of errors that happen in the world. I'm just thinking back to, you know, early 2011 or so, when a very large pharmaceutical company posted their financial statements, and there was an Excel formula error. I believe they lost 24% of their stock value in a day.
Now, a company this size that I'm thinking of, they can weather that storm. But maybe some of your clients can. And really, did they need to weather the storm? Probably not. So I think there's lots of ways that leveraging a technology like this and supporting a group like yourselves, as professionals, with years of history, that are, you know, able to execute good judgment being fundamentally part of the accounting process and understanding of it, partnered with a tool that was designed to help in that same vein, I think we could create something really quite interesting.
So we're getting kind of close to the top of the hour, and I do like leaving a lot of time for question and answer. I think, the things I would say are next steps, before I go into the Q&A, is keep learning and exploring. I believe that artificial intelligence and machine learning capabilities are absolutely in your environments today, in your client's environments today, but this is something that's going to really drive additional success for you.
So learn about it. Be more explorative. Try more of it out inside how you can change your methodologies and processes. Listen, rally the troops. Create a plan and execute the plan. And obviously, if you've liked what you've heard so far, I couldn't go into all the different, you know, pieces and functionality and all the different types of analysis we do. This was really to wet your appetite.
Come and visit us. Spend some time with us. So we'll just pause for one second, and I'll start looking through the question and answer list, and I'll start going through and give you some context of what's going on. So I think a great question is, from the professional field today, there's lots of different working paper and engagement type solutions.
And we think those are very valuable in combining all the pieces of information that you've used to gain and refine that audit opinion. And so the direct question is, "Hey, can you integrate to common ones today?" Something like CCH Engagement, or AdvanceFlow from TR, or things like that.
And the short answer is absolutely. So the solution itself today, and I'll just flip back to it for a second and talk a little bit about what we do. So if I go back to the initial parts of just loading data and looking at information, this account mapping, you can import anything, any file that's coming out of most of the mainstream working paper solutions, including engagement, and your grouping and mapping is done.
You can grab the opening and closing trial balances, again, right from those same tools, which you've likely loaded up. Boom, imported, validated against the mapping that you've just given us, all done. Then when you get to ledger details, or the payables details, or the vendor master list, or all these other files, we then use that, compare it against the mapping and the grouping and against the trail balance opening and closing, and then we do the rest of our job.
So 100%, we can absolutely connect into those, and it's really important, do the job once. If you've already spent time on that, make sure it gets done. So the next one's a little bit long. I'm going to read through little elements of it. So the question is really around risk assessment, replacing current practices, making sure that we can get all the way through the peer reviews and the various controls around our procedures.
And so I think this is actually a really great question because it's kind of hitting probably the biggest stumbling block that some firms will run into, and probably why I say, you know, "Build the plan, and help us help you with that." Because what we do is we actually try to work with you to understand where in the process we can find efficiencies.
We already know from our over 170 clients that we will save you time in the data prep stage, we will save you time in the initial analytics phase, planning phase. We will be able to make sure that we can see all of the pieces. So we will reduce some of those effort points as they are today.
The change management process, and we've got a team of success managers that focus on this with you, will then work with you to identify what procedures we can actually in fact augment, you can read that as a replacement, and be able to actually go through that. So it's not a very, you know, it's not a short answer.
There's a lot more to it. But we do, absolutely, start taking away some of those pains. And what you'll notice is, even as early as late last fall, you've got various standards bodies putting out various papers about how to start thinking about incorporating AI. Our team is there to help you.
Our team is there to get rid of. If I was to take a guess how much overall time you're going to gain that, it's probably going to be in the neighborhood of 10% to 15% minimum based on all of our client engagements done today. You know, that begs the next question, which is from Jeff, thank you for this one, which is all about cleanliness of data.
So at the end of the day, data is a messy, yucky thing that we get from our clients, and it's sometimes more painful to go through as you would want. So two statements, the first one is it's actually not that difficult. And what I'm going to do to demonstrate this is I'm actually just going to go and add another GL data set. I'll just say, "New test," and I'll load in a different data set than the last one I loaded in.
So let me go and grab something that needs a bit of work. So the first thing the system does as it's going through this is it actually says, "Oh, I need a couple of things to make this work. I need to understand if you've done an amount field or whether the data was reporting based on debit and credit." And what ends up happening is you have all the ability to just go ahead and drag and drop fields, if there are fields that we don't understand right out of the box.
We've been training this part of our smart ingestion for a little over 24 months, and we constantly get better at finding new ways to interpret the data. If we are not sure, you're going to get a mapping view like this. Once you've done that mapping, the system will take care of the rest, that will do all the other normalizations and the like.
Same thing with pulling in and verifying the accounts. If there are any accounts that look to be odd or we're not sure where to place them, you'll get a little screen and you can do that. So it doesn't have to be clean. Obviously, if you do a direct connection to QuickBooks, or you grab a QuickBooks desktop file or a backup file, or you grab something from Xero, those known ERPs that are not configured in a strange way, guess what, we're going to be able to go through and just load them up.
If they are clunky, someone's made a change to our report, it's going to go ahead and it's going to say, "We don't know about this or that,"and you'll be able to do that. And the last thing I'll point out is, every time we do process a file, we actually give you the full details of what we thought the file looked like. So, in this case, we're telling you who loaded it, when they loaded it.
We're showing you, obviously, any rows that were, if it's a group report, any that were header rows. So you can actually validate against the file that your client gave you and make sure we didn't miss any entry or any transaction. So I hope that helps with that one. You know, I sort of kind of get this question in different ways.
So this one's kind of funny. For me, you know, it seems easy enough that people could just pick it up and use it straight out of the box. So I think that's part number one. And part number two, I'll get into in a second. Yes, the goal from our perspective is accountants went to school to understand accounting. They got their certification because they wanted to help businesses. And they do not need to be technologists.
We've built something that we believe can be just that easy to use. Clearly, you're going to want to understand our control points a little bit deeper than we were able to do in this webinar. Clearly, you're going to want to spend a couple more minutes just understanding all the other pieces in terms of just getting to the audit plan and being able to do your work.
But our onboarding process, our training ramp-up program, our videos that are available to you, we can get that done in hours easily, for anybody. But the second question or the second part of this, is why I was chuckling, is, well, does that mean that we're going to have to recruit different people? And I think the answer to that one is a little bit more complicated and complex.
I think the short answer is yes, and if you look at your...let's just use the top eight or nine networks in the world, including the big four, and they're already doing this. You see statements a couple of months ago in "Accounting Today," where one of the top 10 networks has hired a chief data science officer.
You can see posting on many of their, you know, job boards that they're looking for analytics people, they're looking for data science type people. And so, yes, I think that recruiting will change a little bit. I think this actually gives you the opportunity to double down and get accountants, folks that are going through the accounting and business programs actually re-interested and re-invigorated in accounting, because they don't have to use just Excel or some archaic tool where you need a bunch of scripting.
They can just log in, grab data from their client, and start, you know, working through what we think are some high-value things. So I think recruitment may change. I think we've probably got time for just one last one, and then anything else that is in here, we'll try to get to as we send out the recording. You know, when we think about the regulators, when we think about the various standards around the world, what is it about those that are going to change, or what is it about those that will, you know, how will we fit in, I guess?
You know, it's like, how will we change procedures? How will we change control testing? You know, what substantive testing can be reduced? All of these are, you know, fascinating and great questions, and all part of the same thing, which is I think, to say, MindBridge will absolutely have a place as things move forward.
Depending on what jurisdiction you're in, there's already statements around how to essentially embrace AI, embrace the ability for free to use a risk-oriented view of the data as part of your sample selection and part of reducing your sample size. And I think what we're doing very, very much, I was with most of the major bodies just this week talking about an education program, talking about how MindBridge can actually help them in, you know, working with the membership around the world, and I think it's a big task.
But I think over the next, you know, you can see it in the statement made by the AICPA president about things changing so quickly, I think you will see, over the next 24 to 60 months, massive changes in statements and guidance. So let me leave you with a really good news, right.
The really good news is that MindBridge has been around for quite some time. We've been training the system for the last three years to help you. We came out to market about a year ago. We've performed thousands of audits with your compatriots and other firms. And we're just getting better, and we're just getting started. There's that much more to do, but it's a great place, where every time you use it, you're going to find that much more efficiency.
Think of this as a subscription there to help you, and it's a solution that you can use on pretty much every engagement and find something useful and interesting. So I know we're right at the top of the hour, being respectful for your time. I know that my colleagues in my marketing group and in my sales group have a couple of other questions that I see just popping in late breaking.
We will get back to you on those. We'll send you some more information. And when we post the recording, feel free to come back to it. And at any point in time, come visit us at mindbridge.ai, from a web perspective. You'll find lots of resources, videos, all sorts of different things and also upcoming events. And feel free to chat with the success agents that are sitting. On the little bottom right-hand corner, you'll see them blue and it pops up.
Feel free to chat with them. They are real live people, not bots, you know, working away, trying to help you get a little bit further on. I want to thank you again for your time, whether it's morning, noon, or night. I really do appreciate it and hope you all have a great day. I'll leave the bridge open, just in case anyone wants to post questions.
And then, obviously, if you do, we will get back to you very, very soon. Thanks, again, and have a wonderful rest of your day.
MindBridge Ai Technology, Mission and Product Overview
Learn about our MindBridge Ai technology, why we created it and how we take a hybrid approach cross-correlating business rules, statistical methods and machine learning to deliver better financial analysis.
Interview with Becky Shields of Kingston Smith
Kingston Smith, a top 20 UK chartered accountancy and audit firm is leveraging MindBridge’s Ai Auditor. Becky Shields talks about the value that Ai Auditor brings to the firm.
How AI is Disrupting Audit and Introduction to Ai Auditor
Learn about the history of AI and why it is emerging now and how AI is currently disrupting audit, followed by a demonstration of the Ai Auditor platform.