A legal accountant explains how AI and machine learning are transforming bookkeeping by reading documents and spotting financial patterns faster than humans ever could. Tools like Clio Duo can surface hidden client trends and forecast revenue, but they still need human review to catch mistakes. The speaker recommends secure paid platforms and treating AI as a tool that enhances, not replaces, human judgment.
Chapters
This week we're going to discuss the role of machine learning in legal accounting and how it can be a game changer for bookkeepers. And specifically, we're gonna talk a little bit about the software that we use. So I'm someone who spent quite a few years knee-deep in legal accounting, and I can tell you that machine learning is rewriting the game rules. It's really gonna make a big wave in our type of work that we do, kind of like Hubdoc did a long time ago when we used to type everything in manually and Hubdoc came along and I'm like, really? It can read everything that's there? And that's what we're looking at. So AI or artificial intelligence, it's not a fancy buzzword. It's not something that's gonna be here today to go on tomorrow. It's a powerhouse where the data can be taught and it can learn and grow based on what you tell it to. So that's the fun part about it. And hopefully we won't really be reliant so much on spreadsheets. Always gonna have our gut instinct. But let me just let you know why I think machine learning is gonna be a big deal. So I think if you really stop and look at it, at its core, machine learning can really analyze massive piles of financial data. This is where it's big in Clio. Clio Duo is their product. And it's really a way to, if you're searching for something, if you've ever had to search in a Google Drive, you know what I'm talking about. Where's that document? Where did I leave it? So it's gonna have the ability to really do a massive search very quickly. And it's gonna be able to take the next step to the next level. It'll be able to search for patterns, look for trends. So if you wanted to analyze a financial statement, you're gonna be able to throw that in and have it like expand the dates and have it look at financial trends and be able to do it a lot faster than our, our, the way we're doing it now and using our brain and our knowledge and our, and, and all the stuff that we've learned over the years and honed our skills. It's gonna be able to do that a little bit fa— do that faster, much faster. And it will constantly be refining itself and growing and learning as you teach it. But it's gonna give you a different level of precision. You're really gonna be able to slice and dice the data, and that's really gonna be a big win for our client and also for our, even our own firm. So we're gonna be able to do predictive analytics. And what are those? That's kind of like forecasting, right? We're gonna be able to see into the future, look at today, and it's really one of the things that machine learning is gonna assist us with. It's gonna help us learn by looking at what we have, what, what it could possibly see in the future. And you could run different scenarios and help it pick up forecasting trends and look at maybe the next quarter's revenue and see if you think it's gonna dip or it's gonna soar and how are we gonna manage it or manipulate it or work through it? What kind of financial strategy can we come up with? This is that's where it's gonna really be able to help us a lot. Now, I don't think— one thing I have to say that, that I worry about is I worry about the fact that we don't have to think anymore. For example, really easy example, I'm a yoga teacher. I haven't taught in a long time. I'm a little rusty on my skills. And I started to kind of look at my old books and look at some old sequences that I put together. And I went to, to the AI and I asked it to put a sequence together based on traditional Chinese medicine, the current element, which is wood for spring, and theme the class around that and give me an awesome quote for the end of class. And it did that within seconds. Now, in the past, I would have taken time. I would have looked at all the different postures, maybe figured out and laid out. And it actually did a really good job of laying out a class that I could use with different poses. And I kind of rearranged them because I like my sequencing a certain way. I don't want people have to get up and shuffle around. So I looked at the sequence and kind of had to still use my brain and my yogi skills to figure out a way to make the class flow better. But I didn't have to think really hard about it like I used to in the past. And I think that that's an example of how it reached out. It knew what time of the year it was. It looked into Chinese medicine and looked at the, the tradition around it and what the element was, where the meridians were, and it did all that really quickly. And I was able to sequence a class like maybe a 10th of the time I used to be able to do it before AI was even around. So I took a break from teaching for a bit. Well, so That was pretty cool, but I still needed to know how to sequence it. And I think the same thing is gonna be the case when it comes to, uh, doing any of our work. Yes, it's gonna help. Yes, it's gonna speed things up. Yes, it's gonna leave us the ability to do things at a higher level, but we're also still gonna have to be, be there with our brain and our background and our history and our knowledge, our wide array of knowledge from all our clients to help with the strategy and look at what it lays out for a plan and whether we will think that plan works and we're gonna have to tweak take it. Much like right now, most people are just using AI to help it rewrite an email, and then you still grab it and you still, if you're, especially if you're taking it and asking it to create the email at first or dropping the message from the client saying, tell me a great answer for this, you still have to take it and fix it. It doesn't, it's not plug and play. Someday it may be, but right now it's still a little rough around the edges, and you want to be able to, to be able to track that and figure out what's the best practice there? And I think with our clients, you can imagine that if we use AI for looking at patterns of payments from the clients very quickly, you're already seeing this in QuickBooks where it's letting you know that this client on average pays their bill over 30 days, over 45 days. That's gonna give you that. That's right in the page of the customer. If you haven't seen it yet, you will. And it's very eye-opening. It really puts it in your face that it arms you with the knowledge that this client takes a long time to pay you. And then you can look at how maybe you wanna change the way that you accept payment. Maybe you wanna take payment before you do the work. Some of those things are, it's gonna take a lot of the grind out of the work that we do, and it's also gonna enlighten us in some details or some trends that we never really thought about looking at. You might know that in the back of the head, your head, that, oh, this client, I'm gonna bill them today, but they're not gonna pay me for 2 months. And you're aware of it, but also seeing it in the data is really helpful. So I think you're gonna see more automation, you're gonna see more accuracy. I think it's gonna be more than just about the numbers as we start to add and work in the AI in the way it's intended. So yes, AI is— they swallowed the internet. So all that data from the internet is put into this machine, and now it's working through regenerative and the learning, and it's growing, and it's been able to— it's able to analyze and assess at a different level, in a much faster level. And you're gonna see that there's gonna be a lot of ways that we can save time without having the errors. And really reduce error rates and breeze through things really quickly and accurately, which is the best part. We're gonna get more time in our day, much like with Hubdoc did when they started reading the data. We weren't typing every detail in on a bill. It was just like zip, and then it attached the document. I remember being just blown away by that technology, and that seems so silly now, but this is what we're looking at as it comes into being working with legal bookkeepers. It's really— our job is more than just balancing the books. It's about keeping the firms thriving. Machine learning, machine learning is really going to hand us the tools to make our job better. And I think if we approach AI and look at it that way, I think you're going to see as you start to test this software out, help train it, help it learn, make sure you're paying for it, make sure you're in a secure platform that you're using with AI. Don't just use the free version, even if you're a small firm. Mm-hmm. Please pay for it, get those security protocols in place, and I'm gonna guarantee you're gonna see some boost in profits. And the best part of all this is you're gonna be able to deliver a better product for your client. And who doesn't want that? So I hope this was helpful. I hope this gets you excited about AI. And if you wanna learn more, we do AI every month in the Law Lab. We share what we know, we share what we're trying. We're kind of really on the cutting edge of as this technology develops. So join us and we'll see you all next week. Bye now. Bye.
