How to Implement Machine Learning in Legal Accounting

This week I'm going to keep on the topic of AI and how do we implement machine learning in legal accounting. So a little bit different than the artificial intelligence, the regenerative type of AI. So when we talk about machine learning, maybe you're a little curious or you're maybe wondering, well, how do I make this happen? And you can see it I think the best example to see how machine learning works is when you see a— just look at the bank feeds in Intuit's world. So they've purchased a lot of AI companies and they are trying to implement the product. If you look at the way the bank feed works, you can see that they're trying to help you make work easier, right? So the long run, we should be able to log in and not have to go into every single transaction every month and put them in. You should be able to let the machine learning start to learn your patterns, your habits. That's really what we're talking about. So when you jump into some AI-powered tools, it can feel daunting, especially when you're used to using tried and true ways of managing your finances at the law firm. So if you are doing bookkeeping for a law firm, you might notice that you can work with firms that maybe perhaps don't do this. Actually picked up a client this recently that They— the bookkeeper was older. She is retiring and she did not want the bank feeds turned on. She thought it was a distraction. She found it to be annoying. She also had everybody in the firm print their paychecks. These are the absurd things that you run into that you think about, how can that be possible? How can in today's world a bookkeeper shut off all of the things that really are future forward, future ready? And to me, that's— it's astonishing when you look at some workflows and you'll find this as you work with attorneys, you're going to find that they're sometimes, they're very paper-centric. They're still back in the dark ages of printing and printing checks and depositing them into the, from printing checks from IOLTA, depositing them into the operating. We still see a lot of that. It's crazy. 2025 is crazy. So for you to start to play with this, I think the best way to start to really test AI to see how you can use it is the same way we've done it for every single tool that we've implemented at our firm anyway, and how I've always believed to do it, is you want to start with your own work. Your own books, your own stuff, your own business, just start with setting your needs and goals. So take a hard look at how you do your own work. If you're like most of us, we always leave our accounting for the last bit. I try to stay on top of ours. I try to do it monthly and pretty much am very successful at that. I make it a rule, hard and fast rule, don't let our stuff get behind. But maybe you haven't looked at maybe predicting your own cash flow. How can you do that? Maybe you look at where you want the machine learning to start to help you solve. Maybe start to work with some predictive analytics or forecasting in your own firm. It's— there's an ability to do that. QuickBooks has some forecasting in there. Why don't you test that in your own company first, and then compare it to what actually happens to see if, if it works for you, and then you can implement it with other— with a client. You really don't want to implement it with another client, or a client, or multiple clients until you know that it it works. Then if you want, you can start to look at some tools that maybe you might want to work with the attorneys. Not all machine learning tools out there are equal. Some are good, some are not that good. Some are add-ons you can add to QuickBooks. And some even— Xero has some of these things, some AI integration, integrations. But you want to look at them and look at maybe, maybe just using a product to audit your books. Keeper is a great example of that. They can kind of audit your books and maybe pick up— I could foresee Keeper in the future highlighting areas that need to be addressed without you having to look through that litany of data that we look at now. There's certainly some products out there that can do that. It can analyze and figure out if things are in the right places so that you don't have to manually do that. It's a task, it's a tedious task to do that. So research the right tools. And then the next step really is getting buy-in from your team. So once you've figured out what you wanna use for clients, getting buy-in from the team to start to implement this. You have to have them on board to be able to perhaps really— if they're not on board, what are you going to do, right? You need them to be convincing to the client to be able to get that to happen. And maybe just discuss with them, you know, the predictive analytics can help spot things which we already see this in QuickBooks. It'll spot slow paying— yes, slow paying clients. It's in QuickBooks now. You can see it in the AR section. Start to play with that. Start to look at that. I've— we've never had that right in front of our face. It's starting to look at the history, the patterns. And typically this client takes X amount of time to pay you and keep that in mind as you're starting to go forward and you'll have that information to share with the client. And then start with maybe a smaller client that's open to you playing around the tools and is open with— maybe once you've tested on your own firm, testing it with their firm. You don't charge them a lot, or maybe you just slash the time just to see how this will work. There's plenty of firms out there that even have AI tools that can actually do a lot of the booking of the transactions so you don't have to. Yes, we have bank rules, but there's actually companies out there that you can hire that will attach their software to your, your dataset, and then you just start to train it to put things in the right categories. If you start to see that it wants to put them in certain places, I mean, the bank feeds aren't perfect. I just had a client the other day, I went in and I could see that she had actually downloaded all the things that Intuit suggested, like uncategorized expense or uncategorized income. So that's not a place where you want to see data. This is going to be a higher level of that where you start to see that and then you start to train it to go where it's supposed to go. Train yourself, train the team. You want to start to really use these tools. If you don't use these tools, you're going to be left behind in the wonderful world of staying ahead of technology, which we've been running through pretty much half of my career is staying on top of technology. So you want to start to use these tools, start to integrate them, scale up your firm, get it to work in a better, smoother way. Some of these tasks might be tasks your, your, your staff do right now that you can now level them up to the next level. But make sure you monitor them, make sure you tweak them, and definitely celebrate the fact that you're not having to do the things. I still can remember this day celebrating with my daughter that I no longer have to type in all these things with Hubdoc. When it magically put things in, and that was an amazing time, and that was many years ago. So when you look at technology and where you come from and where you're moving to, it's very exciting. But we wanna take it in little baby steps, and that's my best advice. So I hope this was helpful, and until next week, bye now. Bye.