Crypto Accounting Firm Uses AI to Extract Blockchain Data Into Workflows

Heat. Hey, Heat. Welcome. We'll get started shortly. Before we begin, here's how to earn continuing professional education credit for attending this session. There are no polling questions or attendance checks. Instead, you will demonstrate your knowledge by taking a brief multiplechoice quiz using the earark app. Visit earmark.app in your web browser. You can also download the mobile app by scanning the QR code on the screen. Log in or if you're new, create your free account. Be sure to use a valid email address because that's where we'll send your course completion certificates. From the home screen, scroll down to find the course. You can also tap the magnifying glass icon to search for the course. Tap on the course artwork, then the enroll button. On the course tab, you'll find the recording and quiz questions. Complete the quiz to earn your CPE certificate. It takes us a few days to create the course on the app. So, if you are attending this session live, please wait a few days. We'll email you a link to the course when it is ready. If you have questions or need help, email us at support earmarkcpe.com. And now on to the event. Hey everyone, welcome back to Earark. I'm Blake Oliver. I'm really excited about today's topic. We are going to see how a crypto accounting firm is using AI in daily practice. My guest today is McKenzie Thod of Hash Basis and she's doing something that I haven't seen a lot of accountants do which is building custom tools to pull blockchain data straight into workflows without writing any code. So I am so happy to welcome McKenzie thoda to earmark. McKenzie great to have you here.

Hi Blake, thanks for having me on. I'm so excited to share all the cool tools that I'm building.

We connected a few years ago when you were building your crypto accounting practice and that was, you know, before the AI era.

And I remember when we talked about this

about crypto accounting in particular, how difficult it was to like get the transactional data that you needed from the blockchain. Like it it took a lot of tools, a lot of like working with data in Excel, massaging it. And I'm really curious to see today how AI has helped you to streamline all of that.

Yeah, it definitely has. And that's been the biggest headache of being a crypto accountant since I started doing this back in 2021 is just getting the data. Everyone always says everything's on the blockchain, which it technically is, but actually getting the data and getting in a format that you can put it into the GL has always been the most difficult part of my job. But now it's it's kind of turned on its head where now I can just add a prompt to an AI and it can build an app, connect to an API, and get me the data that I need. Um, now it's not as simple as that, but but sometimes it is. So yeah, things have changed a lot.

So I'll hand things over to you and you can uh show us what you're building. Um, yeah, where do we start? Like what is what is what is the the the thing we should know first?

Sure. So let me actually maybe start with the agenda. So that might be a good place to start. Um the way that I've broken this down is that the first three sections are more about how we're using AI in our practice in a more I guess standard um way. Like we still do bookkeeping, we still we still do tax. We still have to communicate with our clients. And so I just want to show you guys how we're using AI for these more traditional aspects of running a firm. And then the last part is going to be that's my favorite section is how we're actually using AI for our crypto accounting workflows. But I still think it's I still think it's useful to show how we're doing AI just for like big reconciliations and and normal fiat um things. So we're going to cover that. And then last, we do have a section on data security because of course accountants that's always top of mind is we have all this sensitive um client information. How do we make sure the AI doesn't take it and run wild with it and do things that we that we don't want it to do with that information? So that's what we're going to cover today. Um, but I guess to to frame this conversation, I feel like my mind shift has really changed u mainly since the beginning of this year because that's when claude cohort came out and all these tools just kind of burst onto the accounting scene. I feel like the way I think about my job has has changed a lot. Um, and this is more so this applies more to I guess the the crypto aspects of it. But my old mindset was kind of I really hope I can find this data somewhere. So one of more so of like desperation I suppose um because again getting the the blockchain data was so difficult and if I couldn't easily get the data like through a crypto subleddger or something then I'd have to go through so many different avenues. Um and when I first started crypto accounting we just had Google so there was no I couldn't even ask like a chat GPT or any kind of AI chatbot to help me. It was just digging through a bunch of links maybe going on GitHub like trying my best to be a programmer. And so, yeah, I guess my mindset was, oh, I might put in all this work and I might not even get what I'm looking for. So, I I should say it was a bit more more passive, I suppose. Um, but now my mind shift was totally changed to I define what I need and then the AI just goes out and the agents just build it for me. So, now I just feel a lot more empowered. I feel like like an architect and I I tell the AI like, "Here's exactly what I need. Here's the format I need it in. now just go do what you need to do to to get me what I need to upload into my like into QuickBooks or something. And so it's just a lot more it's more fun to be an accountant these days I think.

So you're basically building apps and

like did you ever study like computer programming? Do you know how to code?

Um I tried when I first got into crypto accounting. I did take some Python courses uh and maybe some like JavaScript as well but it wasn't it wasn't that useful. fly because I was doing it on Code Academy, so I had some fundamentals which were good, but it was I really struggled to to implement it in my day-to-day work because I'm still kind of scared of like using the terminal and things like that. So, it was just it was just never going to happen for me. But now it's like I don't need to code. Like I'm kind of glad I didn't do a boot camp and spend all this time learning how to program because you just you just don't need to. At least at least for the level of apps I'm building is they're not like super complicated with like backends and stuff. So vibe coding is totally fine.

That's amazing.

Yeah. So that's been my mindset change over the past year. Um, and this first first section that we're going to cover is how do we enhance our bookkeeping that we do at Hash Basis uh with AI. Now, just to set the the tone here, like AI is really good at some things and then other things it's not so good at just yet. Um, and I think it's good to keep in mind that we're still really early, not only in the AI world, but like using AI in accounting. And so things are going to evolve. And I'm super curious like next year if we like have the same conversation, how are things going to be going to be different? So this is more so like what we're using in our business right now. And just a couple of like high level things of what works, what doesn't. Um we find that reconciliations AI is really good at like you and I were just talking about this. So things like uploading your bank statements and connecting your AI to uh like QuickBooks via an MCP or connecting to your bank via an MCP and then the AI can just build a reconciliation for you or or do the wreck for you. And we're actually going to look um an example of that later. So we've been using that a lot. Um skills, which I I think most accountants are somewhat familiar with this, but basically it's just a set of instructions that the AI you can like save to the AI and it remembers how to do tasks um for for future reference. So that's really great. As an accountant, most of what we do is just kind of like a repeatable task workflow every month. And so you can just do it once, tell the AI like, "Hey, save this as a skill." And then now it knows how to do it um more efficiently for next time. So AI is great at that. And then browser automation. Um this one's really cool, I think. So what the AI can do is just spin up like a like a Chromium browser in your tab and then just click on your behalf. And so it just mimics you like clicking like add a customer or approve all these bills that I have in Netswuite or something and the AI can just I don't know how it does it on the back end but it can just click for you which um accountants know a lot of our job is just mundane clicking like the amount of carpal tunnel I've had over the years is crazy but now it's like basically gone because I don't have to do this like incessant clicking all the time so AI is super good at that. Um now there's a couple areas I noticed where it still kind of falls short though. Um, one of them is that these MCP connectors, which basically it lets you connect the AI to your various applications, a lot of them aren't very good yet. Like QuickBooks, the big one everyone's talking about now is that you can't actually like push journal entry to QuickBooks via the MCP connector. And for me, like I I post entries a lot of the time. So, not being able to do that, that's just kind of annoying. So that's more so like I think that's a QuickBooks thing like they need to be able to expose more actions to the user like for the AI to do. So that that's kind of a drag right now. Um the browser speed when you spin up that virtual browser it's really slow and so while it's cool to look at um it's not actually like a person doing it would would still be faster. So speed is definitely a consideration. So that's wh