Sodium, William AI and Xero Claude Integration, QuickBooks MTD Checker, Combinely and FreeAgent Ecommerce

[00:00:00] Hello everyone and welcome to the Digi-Tools In Accrual World podcast, the place to find the latest accounting tech news. I am in the chair today running the podcast and with me I've got Lara Menton and Robbie White. Lara, how are you doing? Good, thank you. Nice to be on again. Yeah, it's good to have you back and you too, Robbie. Oh, it's an absolute pleasure, Lee. I'm super excited to talk about a bunch of stuff today. It's going to be a whole lot of fun. Yeah, we have got a lot of stuff to talk about today and we will try to talk about some non-AI stuff because that seems to be everything everybody wants to do at the moment.

[00:00:29] We've got a lot of AI stories but we've got some stuff from Sodium. Duane has kind of released Sodium, a new practice management tool which we'll get into. You've got some stuff on Apron as well to go into, some more AI there. Yeah, their release of, the next sort of release of Will, William, Will-I-Am or Will-IA or whatever we're calling it nowadays. But yeah, William the AI, their AI robot. Yeah, definitely. Robbie, what's your top story from this week?

[00:00:57] There's so many but I mean I think that the two I'm really excited about, if I'm completely honest, is some of the amazing work with Clawed and its integration into Zero. Like this sounds super exciting and then also, personally, I like some of the web-based stuff that Combine Me has released. So, in the browser, trying to make things easy for accountants, like that's, you know, that's where the future of this all comes down to. So, I'm super excited. Will say it's on AI though, so unfortunately, Lee, we are going to go down that rabbit hole again. Definitely.

[00:01:25] And we're just coming down off the whirlwind of account techs, aren't we? Were you there, Robbie? I wasn't. I'm so bleak. I missed out on what sounds like an amazing time. Give me all the news, guys. What's the, I guess, what's come out? What did you see? Anything that stood out specifically? I'd love to hear. I think it was quite interesting. There were a lot of new vendors there that were doing quite a lot of the same things. Obviously, a lot of AI around it. There wasn't anything that really massively jumped out.

[00:01:52] I spoke to a few people and said, what was the biggest new thing that you saw there? Or what was the next big thing? And they were like, not quite sure what that was. There wasn't like anything that massively jumped in your face to say, this is what it, this is the next big thing. But lots of different uses of AI, different smaller tech coming through, whether it's content creation, just management, document capture.

[00:02:17] And it's quite interesting to see the apps that were sort of smaller a couple of years ago now getting bigger and bigger. So that was my main takeaway. As you can see who has come up kind of through their size stand, I guess. So it started here and then we kind of got bigger or more positioning. But yeah, it was pretty busy two days for me because I spoke four times. So I didn't get to see as many people as I thought. Definitely. I think the other thing with the stand size is maybe who's raised money as well, rather than who's necessarily selling a lot of stuff. Right. So brilliant.

[00:02:47] So lots of stuff for us to get through. Let's move on into app news and start getting into these stories. We're welcoming Fishbowl to our podcast. Fishbowl is inventory and manufacturing software for businesses that have outgrown spreadies and bolted on add-ons, but aren't ready for full ERP. They've got over 25 years building this product in the US and 15 years in Australia. And now they're showing up properly in the UK, both desktop and cloud working alongside QuickBooks and Xero.

[00:03:16] What stands out is their manufacturing depth. There's multi-stage build of materials with sub-assemblies, mid-build component swaps without having to cancel the work order and reissue it, and capacity and production planning. True landed cost reconciliation and all four costing methods, VIFO, LIFO, standard and average, which is rarer in this category than you'd think. The other thing worth knowing, when you're quoted, the price includes the product.

[00:03:41] So barcoding AI insights, the reporting modules is all in. They're not on a list of extras. And when you call support, you get a lovely Fishbowl employee. So the accountability always stays in-house. It's worth a look if your inventory tool wasn't built for the operations you're actually running right now. So I'm kicking off app news this time with the release of Sodium now generally available. So Sodium is a practice management tool.

[00:04:08] It's built to incorporate AI from the ground up, which I think is a bit of an advantage for newer apps that are coming out at the moment. So rather than have to re-architect or try and bolt some AI on, you can kind of build it from the core at the beginning, which is really interesting. All of the sort of points about the release and the story and stuff are in the app newsletter, which you can subscribe to and go and find the points there. But the thing that really jumps out for me on this is it's a very, very crowded space, isn't it? The practice tools market at the moment, loads and loads of new stuff in there.

[00:04:37] You've got the sort of incumbents with their big suites. You've got even, I think, Rachel Harris announced a new product at AccountEx as well. And I think the challenge for vendors in this is the sort of chaos that goes on in firms around this stuff. And the best way I could think to describe it is if you gave a toddler a load of Lego and they just stuck it together and made a little wall, it's all jumbled and it's all got different sized pieces and different colors and stuff. And that's how this stuff tends to work in firms. People have just sort of got pieces together and stitched them in.

[00:05:04] And so the real challenge with a new market entry is they might come out with a little golden Lego brick. But the challenge for the firm is, well, what bricks are you going to take out? And how much do you break that wall apart to put this new bit in? That makes these things a really, really big IT project. But at a level where you're only paying a couple hundred quid a month, so that maybe the enablement support isn't there from the vendor and it maybe isn't taken as seriously as it should be. And so that will be a real blocker, I think, to break through for these apps to get traction in this space. What do you guys think on that?

[00:05:33] Well, I don't know about you, Robbie, but I know myself having moved practice management a couple of years ago, what a headache it is. You've got to be really sure that it's going to give you what you want because you've not only got to get yourself moved across, you've got to get your clients moved across if it's got client portals and document management and that side of things. So it can be one where we kind of sometimes we stick with something longer than we should because we're just scared of that move.

[00:06:00] But equally, you've got to be 100% sure that you have the right app in place if you're going to move because it is going to take hours of your internal time. As you say, you might not get that support elsewhere as well, especially with new apps. It's worth, for me, practice management, it's worth seeing what else they connect up to, how else they fit, as you said, with that Lego wall because it doesn't just impact that one tool, it's going to impact your whole workflow as a practice. Yeah, I mean, I completely agree.

[00:06:28] I think, you know, a lot of the, I guess, as you move any technology, there's always the pain of going through that cleanup and making sure it's accurate. I think when it's your core systems of the business, when it's practice management, obviously this is just huge change and nobody wants to undertake huge change for the sake of it. And frankly, it always takes time. Like I've never heard of these things in any practice management solution being implemented without change, change the way you operate, change the way you do things. So fundamentally that's key.

[00:06:56] But I do think as we see these advances, what we will see more of is these practice management solutions getting, becomes a crowded space, but actually the ability to, to change from one practice management solution, I think is going to become automated to another more and more. And we've already, you can start to see it in the ERP space. There are companies being built, which are automatically moving you from, you know, you're on large ERP system number one, and you need to move to large ERP system number two.

[00:07:25] And there is like a whole bunch of pre-work that's done by AI and agents that make that a lot easier. So one of the big things I think is actually, you know, it's a crowded space, but that actually might apply some pressure on vendors to make real changes. So these big vendors that we've dealt with for a long time in practice management, I mean, you know, there's going to be some pressure on them and that ability to change their models and, you know, the capital that they've spent over time. And to your point, find the golden Lego brick. I think that's going to be, you know, super, super important. Yeah.

[00:07:56] And the thing about Dwayne is he's not, he's not afraid to antagonize some incumbents, right? He's got a great track record with that, with what he's done previously. And although there are, there are market challenges there, Dwayne tends to build really good tech, right? He builds a good product. And I think the other notable thing in here is no VC is involved. So he's got full control, he can do what he wants. He's not being driven to hit X or Y arbitrary goal. So hopefully the firms that work with him might find a slightly different experience. So we'll keep an eye in there and see how it goes.

[00:08:25] So the next bit of news is that Apron have launched fully their William AI solution. So we spoke about it a couple of, last time I was on, I touched on what they were bringing out, but it's a mix of things. So you've got the document chasing, it's going to connect up to emails to go and look for those missing documents to help with that bit.

[00:08:49] You've also got a trained AI or an AI you can train to support with publishing. So native language, helping you actually get those documents in quicker. The good thing is you've got the ability to review what the AI is doing. So it's not just turn it on, set it going. You can teach the system. You can see why it's making the decisions it's making. You can see what the percentage of things is that it is publishing for you.

[00:09:17] You can see the confidence that the system has around publishing. I know there's a couple of other people doing that. So for example, Dex AI Assist is doing something quite similar, but it's pulling this all together to give you a bit more assistance on those documents. It doesn't deal with things like tracking categories yet. There's still a few tweaks that need to be made to be able to kind of turn it on for everyone, I would guess.

[00:09:41] But it's coming a long way to actually helping you hold things more quicker and more effectively, really. I'm really intrigued around how that's changing the way that you operate on a day-to-day basis. What does it practically mean for you? Is that a material time saving? Is this just another add-on that we're seeing into existing ways of working? I guess what's the thought? It's a step up from that auto-publish, isn't it? That's where it really sits.

[00:10:07] So auto-publish was, if it was this exact rule, go and do this. And only if it was every single time. Whereas this is now a bit more of that AI Assist supporting with it. You're able to break things down a bit more. You're able to kind of work out whether you need to turn on auto-publish or not as well. So it is going to be probably smaller time savings at the start while you build confidence in the system and while you teach the AI what the sort of things your clients are doing.

[00:10:34] But I would think that gradually that's going to then build up into something a bit more material. This must be a huge decision at a practice like Mazars level, Robbie. Because if you're in a small firm and you go, well, we'll turn on auto-publish. If it goes wrong, maybe we've got half a day to unwind it. But you must be terrified of switching something like this at the scale of a firm like Mazars. Yeah, well, I think the other side is also with any of these types of toolings, how is it going to impact our clients?

[00:11:02] At the end of the day, where our focus is going to be is ultimately what is the client experience going to be around these things? So, you know, what we really focus towards when we do these rollouts, and that's why I asked, is like what's the actual practical impact that's going to have? So if it's time-saving for us, that's phenomenal. That really helps. But, you know, what we want to do is really make sure that there's more time given towards clients. So I think the biggest thing here is really whatever the technology is, we leverage other, you mentioned DexterSyst and other types of tools in ways of working.

[00:11:32] I think whatever you're doing with the tech and how it optimizes your work stream is amazing. But it's what you do with the extra time and how you spend that. To your point, though, Lee, when you roll the stuff out, I'm sure Laura went through it too. It's not a big bang. We don't go out to everybody kind of small and iterative. And what I will say is more and more clients are asking these types of questions and more and more clients are kind of saying, you know, what's happening out there? Can you give us a perspective? What does the market look like? The challenge is it's very noisy.

[00:12:00] There are lots of, you know, William or Will.I.Am or whatever we're going to, you know, as an example, like these things come into the market. They're super important, but they're one of many. And it's how you, you know, how you position that for a client to make it really, really easy. I think there's a twofold thing because there's a practice bit and then that document chasing bit, which is hopefully going to help the client. That's going to then go and look in their emails very specifically for the things that are missing. And again, that's something where AI can assist the client getting that document to us quicker.

[00:12:30] And I think that's going to be a big thing for practices to have that all in one bit of like you've already got the document chasing within apron, but it's like, oh, actually, you don't have to go find it. William will go find it for you. So that's that next level. It's really interesting is, you know, one of the things I'm going to share a bit about is, I suppose, what's happening on the other side where you've got these really large, like flawed, you know, being implemented inside of Xero.

[00:12:56] And this is an example, you know, where apron is maybe more focused and we've got William, which is a clear task, clear operation. Floored as a specific is way broader. We're talking about something that covers the, could cover the entire ecosystem. The challenge with it right now in the work that they've done with Xero, and I'm sure Lee has lots of opinions on this, but the challenge is right now, it's really early stages. And there is not a whole stack that's in there which has matured.

[00:13:22] There's lots of really interesting things that they can do around cash positions, you know, reporting on where they are, giving clear insights. But actually writing back into Xero, like as far as we're aware and we're still looking into it, like that's not something that they're doing with flawed. And so that optimization that you spoke about, Laura, that's not kind of embedded natively into what they're doing.

[00:13:45] But you can also see that if this is the kind of path that's going on, we're going to have loads of challenge because the speed at which flawed can operate and the amount of people that are probably going to invest time there is huge. So you kind of have this, you know, what do you start to back? Do you back these niche areas where we've got specific tools coming out like an apron or do we start to say, well, actually, you know, there's these massive companies who are investing huge amounts.

[00:14:14] I mean, Claude and Anthropic and OpenAI have both released, you know, deployment teams. They're investing billions around like deployment teams to focus on how they're going to build out these use cases. So there's this real challenge, I think, around these large, big businesses in AI versus these small, nimble areas which aren't frontier firms or frontier companies but have specific knowledge in IP.

[00:14:41] But I don't know what your thoughts are, Lee, because you must have an opinion on this. I think it's really, really interesting. This touches on the next story I'm coming up with in a minute with Intuit. But it feels like the general ledgers are hedging their bets by the looks of it. So you've got zeros, got jacks, and they've got chatbots in there. And, you know, Intuit, they've just come out with a new AI dashboard and embedded AI tools so you can use AI inside those products.

[00:15:06] But then they're also surfacing those products inside the big LLMs, like you said, Robbie. And you sort of wonder if there's such a lot of effort went on in the last 10 years or so for these general ledgers to be the place that people come to. They want you to log into Xero. And, yes, we'll pull in your information from Dex, but you log in here. And it feels like they're seeding that ground very quickly to these LLMs.

[00:15:30] And maybe that's because they know that's the direction travel anyway, that people are going to want to just use these tools connected to stuff. And that's what they're going to interface with. But at the time, they're putting the bots inside their own tools and going out to Claude as well. I'm not really sure which way they're going to go. And it might be, like I've said, they're hedging their bets on it. But what do you think, Lara? It feels a bit to me like they're trying to cover the market. So it's like if you already know Xero or QuickBooks, you're going to use the LLM potentially within the software.

[00:15:57] Or if you're coming at it from an AI-first perspective, I know that, for example, the Claude Small Business Connectors, that has an Intuit connector within it. So they're looking at, okay, what's coming in? You're in the AI already and that's where you're starting. Do you then start to look for your software? Because I think that this is kind of a different move. This is coming up now. Are they thinking about their accounting or are they thinking about their AI first? And then they're going to come to those accounting tools through the AI side.

[00:16:27] So I think it's sort of, as you say, hedging their bets, but also covering the market, I would have said. The one thing that's pretty interesting is the agent building inside of these types of tools. So using these large AI tools, the agent building inside of them is really, really strong. So what we will see, obviously, is these unique agents that we have now coming out. But the ability to build more agents at scale, I think, to your point, Lara, that decision around, do we want to focus on the AI?

[00:16:56] Do we want to focus on the general ledger? Where are we doing this? And where is our investment going? I think it becomes a huge thing. Definitely. And Intuit have got two updates on this that seem to be the same thing. I'm trying to figure out what it is. You can have a look at the actual articles in the newsletter. But they've got the next era for the accounting Intuit suite, which has things like getting instant answers through the chat. But there's an interesting thing which is coming soon, which is comparing data across the client portfolio.

[00:17:25] And that always used to be a really, really nice sort of gold standard of reporting stuff if it could give you benchmarking across your practice. That was always a huge differentiator in the early days for the sort of spotlights and the fathoms of the world that they could do that. But that's coming soon on here. So that will be an interesting workflow, I think, for accountants to start giving a different perspective to their clients on what each of them are doing. And they've got another article, which is, again, then just saying instant answers.

[00:17:49] And I think from what I can see, because the article is quite light, that they're offering this potentially to end clients, which I know makes accountants a bit jittery. Like the idea a client could go in there and go, should I buy a van or should I hire someone? You know, the questions that you really should be asking an expert on whether or not the disclaimers are in there saying, well, you know, we can't give financial advice. People being people, if they can ask it, they will. What do you think on that, Lara? I think that's the worry, isn't it? That people will go in without potentially having all of that information in there.

[00:18:18] That's always the one where I think accountants and bookkeepers, we're quite good. We'll review things. We'll look at whether the P&L looks right, the balance sheet looks right before we give that advice. And I think the AI, it could be that they've got a ton of stuff missing. They go and ask the AI a question. It tells them something based off wrong information. And that's where it really makes me nervous is without knowing what's in there, how does it give that right information? Yes, if you've got a completely clean set of books, then maybe it might give something similar.

[00:18:48] But it's not going to have the context around the client as well. It's not going to know what the client's end goals are, what's coming up, all the things that we do by talking to our clients. The AI is just looking at a set of numbers and giving you a boilerplate answer, basically. And if those numbers aren't right or if it doesn't understand the business, that's when you're going to start coming into issues. Just to dig into what you said there, Laura, about that human intuition, because the AI has the context that you give it.

[00:19:13] And in the last episode, we were talking about how there's been some instances of AI drawing from old articles from GovUK and therefore giving wrong information. But then the other thing, I think, from the human thing that isn't replaceable by AI is, let's say you've got a client and they're in their mid-twenties and they've just got married and they've got a single director. You would probably think so. Probably don't take all your income from dividends. Probably get some decent payroll in.

[00:19:43] And AI would never think to ask that because it just doesn't have that human insight, does it? Is that not devaluing what we give, but I think it's understanding a bit more around how that differentiator is. What can AI do versus what a human, as you've just said, that human element, leaning more into empathy and understanding and goal setting and what a client wants from their life, not just from their business.

[00:20:11] I think that's where we've now got to go as bookkeepers and accountants, is to lean more onto that side and say, right, the AI can deal with the rule setting and the things that are every day. We're going to be looking at the bigger picture. Do you think the one thing I'd highlight, though, is, you know, we, whilst that's 100% true, AI learns and this is the one thing we're going to have to start to get relatively comfortable with, is that these skills and these things that we know now, at some point, AI will learn that.

[00:20:39] AI will learn. We've had loads and loads of people that are directors that are mid-20s and are making decisions on their life. Guess what? They all consider the same thing. And as it does that, I think one of the things is, you know, where do we continue to add value in that process? The other part for me is, I do believe that that's more and more of where, you know, our time will be focused, is really continuing to build a relationship. So it's not just, you know, simplified to, you know, making those type of things,

[00:21:09] but it's like we really know where people are. We genuinely know you better than, you know, you're not a number on a large spreadsheet. We've spent enough time with you that we can actually have this type of conversation. So I think that that more and more, we see it across the board of the tech. It will learn. It will get smart. It will take time to do that. But at some point with these humanistic type conversations, I think are going to get deeper and deeper and deeper.

[00:21:34] Do you think then there's a bit of a rebranding exercise that needs to go on for accountants? So they're not just seen. And I know this has been, it's an old tide cliche, and I know that it's not true, but there is a large part of the population that see accountants just in number crunches. And, you know, they don't really recognize and understand that that human element is there. Do you think there's an impetus to promote that more so people don't feel that they can just quickly and wrongly replace it with an AI?

[00:22:01] Well, 100%. I think, I mean, we all know this, you know, the best accountants are not those that have the most technical mindset in certain areas. Yes, they might help on certain technical areas, but it's the people that really understand what the problems are that an individual is facing or a business is facing, you know, and they work with them to try and solve that. You know, the people that are making decisions around the future of a company are people at the end of the day.

[00:22:27] And, you know, they'll use lots of information, they'll want lots of insights, but they'll deal with people. And the more that, you know, conversation kind of changes, the better. I do think there is still a brand that is, you know, what is an accountant? And that will definitely need to change. But I do also think that there is just this, you know, general type of challenge that we have where we worked in a traditional, you know, set of ways of working.

[00:22:52] And that's changed over time. So, you know, we used to be very time-based, very time-oriented. We still are in many parts and that drives a certain type of behavior. We used to be very profit-focused, you know, and how we measured profit was on time and the cost of an individual in that time. So those kind of things drive behaviors inside of companies. Now, as we move towards spaces where we're more outcomes-focused, where we are, again, you know, focusing on where we add real value,

[00:23:20] that the type of conversation I just think fundamentally changes. So whether it's rebranding or it's just, you know, trying to be clear on where the value is, I think that's super valuable for the future of the profession. So this is where I think AI is really, really hot at the moment, which is QuickBooks's new MTD checker, which is AI-powered. Now, I know that bookkeepers love their VAT checker, so I'm hoping this is going to be quite similar in that way.

[00:23:47] So apparently QuickBooks had the highest number of cumulative MTD sign-ups during HMRC's testing period from April 2025 to January 2026. So this is something they're obviously being quite proactive on. It's going to flag duplicates, missing transactions, looking at all of these things. And it's also going to support non-trading income sources, making sure that you've got everything in there.

[00:24:13] So this, obviously, we are now in the start of MTD. We're coming up to that first, the first submission shortly. And I guess this is building upon what they've built in product, but obviously assisting that, making sure that for those first sets of checks where everyone's going to be a bit unsure on what goes where, how it all works, what that final pull together is. So if you've been on beta, you're going to be a bit more confident with that.

[00:24:42] But having these second set of eyes on there, as they do, as I say, for their VAP checker, I think it's going to be really, really key. And I think this is a really good use of AI as well, because we're following those set rules. HMRC have obviously given the guidance down. Where things need to go is set. And I think this is one where we can be a bit more confident on how that AI is learning and using. What's your feeling, Robbie, on this?

[00:25:06] I think the biggest thing for me is, you know, well, so first of all, a moment for MTD and MTD like officially being live. And we are now not talking about it being in the future or tools that might help with MTD. It's now practically happening. And we got arrived, live use cases. So for me, amazing, amazing to see this type of automation being built into the process. My challenge on it is, you know, how are people, again, how are people adopting this into their workflows? What does this mean?

[00:25:35] How does, you know, great that we've added in something that should optimize our process. What does that practically do for us? And I think that type of change, you know, is going to continue to happen. We know the tech will be there. We know there will be increased kind of conversations on, you know, what comes out or what doesn't. But for me, I'm really excited that this has come into one of the big general edger spaces. It's being released. It's being adopted. People are using it. We're not, it's not theoretical. So I'm hopeful.

[00:26:04] Maybe I'm just eternally hopeful, to be honest, Laura. But like, I'm really hopeful about where this goes. So yeah, I mean, that would be my general view on it. And I do think the investment from QuickBooks in the space is clearly starting to pay dividends. You know, we've mentioned a bunch of different releases already. Lee's mentioned a couple. In this space. And I think that that shows what the vendors need to be driving towards. So I don't know if you feel the same. I don't know if you feel like they're on the right pathway.

[00:26:30] Just any big concerns, I guess, in terms of gaps that you are immediately seeing with the, well, with the MTD stuff. But more, I suppose, importantly, where QuickBooks are going. I think, as I say, it's a checker. It's those kind of having those checklists, having that second set of eyes, having that support there. So we're still doing the work, which is the key. But then having the AI come in and look at the things that we might be missing, especially at the start while we're building up those processes, I think is very good.

[00:26:58] And having it from day one as well, having it from the very first MTD submission is going to mean that it's going to be a lot easier to embed that in your workflow rather than you build a workflow. Something changes. They bring an AI or they bring in a different version. And you've got to build that into your workflow. Having it from day one, I think, is going to make the key difference to practices. You know, the other side of these large investments is where we're starting to see what people want to get value from the platform.

[00:27:26] So that's where these vendors need to drive towards. And I absolutely love, so one of the areas that I'm kind of speaking on also is around Combinely and what they've done with the, I love the branding was amazing, like a junior accountant in your browser. That's the way to sell something to somebody, right? But that's the principle, I guess, that if we link the two stories, what we're really trying to do is how do we make things easier to adopt, bring them into our workflows? In your case, it's a checker.

[00:27:55] You know, in Combinely's case, I guess they're looking to push that a little bit further and actually optimize and, you know, build agentic workflows into the ways that things are done, which is super exciting. I mean, the opportunity here is massive.

[00:28:09] I think my concern with some of what Combinely is doing is, and if we go back to what we spoke about right at the beginning, you know, if you're looking to build agentic processes and maintain these in the space, at some point, those large vendors are probably going to come in and people are going to start to try and replicate what you're doing. But everything I've heard so far on Combinely, I haven't seen it in action myself, but everything I've heard is, it's super, super intuitive.

[00:28:35] And because it's built into the browser, it actually helps you through all of the different steps that you would be taking on a daily basis. Again, and more importantly, can actually identify some of those steps for you, which supreme help in a time when, frankly, we're just looking for any productivity gains which we can get, I think, as an industry as a whole. So, you know, from my perspective, this is the type of add-on, you know, which is valuable because it's not inside of the solutions.

[00:29:02] It's not baked in like we've spoken about with QuickBooks or some of the work with Jackson Zero and other areas. This is kind of built on top of. It's a new block. I'm going to stick to your story. It's another block we're adding on. But this block, I am pretty sure, you know, the agentic processes it brings with it are valuable. But, I mean, I'd love to hear, have either of you played much with Combinely? I don't know if you've seen much of it, but I've kind of looked on some online demos. It looks exciting. And as I say, the marketing on a junior accountant in your browser is definitely the way to be positioned.

[00:29:32] I've used something non-accounting specific. So, I've used something like Sintra, which, again, is a range of different models bent for different areas. So, it sounds like Combinely are doing something similar with different models that are tailored to different areas, but within your browser that you can send different areas of work to and support you. So, it will be interesting because I think sometimes having that more tailored model and it understands or it has an end goal in it,

[00:30:01] as opposed to just a standard LLM that's obviously pulling from lots of different sources, but maybe not. You've got to teach it a bit more, whereas I think that that's where some of these time savings come, is actually the model having the right direction that it's going to go in from the start, rather than you needing to know exactly how to prompt to get the most out of it. If you've got something that's a sort of a periphery thing like this, it could just sit on the side of a workflow and help. It's easy to put in, but it's really hard to embed it as a habit,

[00:30:30] and therefore it's easy to just drop off and for it to be lost later. And so, I agree with you, it is genius because immediately it gives someone a way to start using it. When it's framed as like a junior in a browser, you're like, okay, cool, I can start, right? It removes that first obstacle. So, the next story is around FreeAgent, and this is an interesting move for FreeAgent. They're going to start doing a bit more around the e-commerce space.

[00:30:53] So, they've connected up with Equally for pulling in from Shopify and different places like that, which, again, FreeAgent to me has always been small sale traders or freelancers, IT consultants, contractors, that side of area. They're obviously building upon that to say, hang on a second, if you've got a small e-com shop or you're selling on these platforms, we feel that we're also going to be able to help. It looks pretty good.

[00:31:22] We're talking about quick setup from the Equally side, reduction in manual effort, and then it's going to pull into FreeAgent, pull all those sales in, have everything in there. Again, I guess around MTD, these are people that are going to be captured, that maybe were doing things here and there, but didn't have everything in one spot or didn't have that general ledger connection. Also, we know with FreeAgent's connection with banks, it can be very low cost or free in a lot of cases.

[00:31:50] So, maybe they're trying to target that area and say, hang on a second, you don't have to have an expensive solution if you are doing a small e-com shop. And that's an area that's always growing, isn't it? E-commerce is one of those things that people jump on and we go through cycles of. But it will be doing, obviously, data reconciliation, categorization, journal entries, sales, all those kind of things, which I think from not just from a bookkeeper or accountant's perspective,

[00:32:19] from a client's perspective, that can be a real headache sometimes, especially with e-commerce and having those things feed in. So, you won't necessarily need one of the bigger e-com platforms or something like an A2X plus zero or whatever it is. This will be designed, I guess, more for, not smaller, but for a certain size of business that potentially can be hitting the MTD threshold, but it's not looking to kind of grow into a global e-com business.

[00:32:45] I'm always a little interested around how people decide to make that decision to move. How do we go and say, okay, this is our core business and now we're going to start to go into e-commerce. One of these areas, I guess, as you logically explained, I suppose, they're trying to grow their business, they're trying to grow their footprint, and ultimately make sense from an end-user perspective or from a client perspective. But I think the biggest thing for me is it's another thing that they're going to have to maintain

[00:33:11] and monitor over time versus others starting to play that kind of platform play. You know, look at what Xero and others have done. They've kind of said, we're a platform and you plug into us. I think what's been really interesting, you know, listening to this is very much it's a, you know, we're going to build the extra parts and add on. Whether it's good or bad, we will only see in the future. But I don't know, Lee, I mean, from experience that you've had before, you know, times at Xero and like this whole platform play and how wide you go with it. When does that stop? Like when does it become enough?

[00:33:40] Or do you just start to, do you just kind of keep adding and, you know, keep maintaining all these different products? It stops when you achieve the average revenue per user that you're targeting. And so if you can't grow through acquisition of new customers, which eventually trails off, right? Like you can have a meteoric rise, but eventually you start to saturate. You then need to grow your revenue in other ways. And so that's what part of this will be as you spread out and add more stuff to your product.

[00:34:09] But a lot of the time it's about revenue growth. And also sort of increasing the area that you can defend from as well. Because if you see a competitor puts a tool in, let's say in this case, like Xero have got stuff around e-commerce, isn't it, have? But if that was a new thing, you might go, oh, no, we need to move into e-commerce. So Xero don't find a way into our customer base via e-commerce, for example. So there might be a strategic reason you'd do it from that perspective as well. I think with Free Agent, they're also actually partnering with Apron.

[00:34:39] That's going to be one of the next integrations. So they are looking to get rid of that kind of, not the prejudice, but what people were seeing around them and move into more of, we are a big player in this space. We can cover all sorts of different businesses. We aren't just for the small, small traders, those freelancers anymore. We can do e-com. We're not just going to rely on our own document capture, or we're going to actually partner with more and more apps in the space.

[00:35:09] So that's, again, an interesting one that they're kind of saying, right, okay, look at us, look what we can do. We can sort of play with the bigger companies in this space. We are kind of, we're now established. They've been around for a long time. They've got a very good record, very good support record and everything around that. And they can then build up to kind of capture some of those businesses, maybe coming in from TD or maybe where the other software has gone up in price. They're looking at, okay, hang on.

[00:35:36] We've got these partnerships with Metal and whoever else to be able to give you this for a very low-cost solution. I mean, the one thing that this does say is that growing out of your space, I think, is going to be super important. Those types of partnerships and where we go. And as a sector, as an industry, I think what we started to see is that this is definitely happening in our space. And I mean, you know, there's been, it was a fantastic talk at the time.

[00:35:58] I still have to catch up on from a context around, you know, how the industry as a whole is starting to kind of bake in what we're seeing in AI and what we've spoken about a lot today into the offerings that I have, into the consulting that they're doing for clients. We will have seen now the kind of big announcements. There's been big announcements from KPMG around their platforms and how they're looking to integrate in AI into those and frankly, all of the other big four.

[00:36:25] But more and more, what we're seeing is that isn't a big four thing. This is like everybody needs to kind of change. And going back to what you were saying, Lee, around the role of the accountant and the brand that we are starting to pick up. I think a big part of that is really how are we changing our narrative from being purely around finance, moving more and more towards being these consultants which provide guidance on AI and also around the decisions. One of the biggest things we've seen is how do we support adoption?

[00:36:54] So I know that we started, we said we wouldn't do loads on AI and, you know, we've spoken loads on AI. I think the other flip side of that is that the ICB is actually putting out some courses to members around the best practice. And use of AI as well to help with that guidance with all of these things coming in and all of these options out there.

[00:37:14] It's making sure that we have that guidance around how to use it, what's the sorts of things we should be doing, what the guidelines are to support us so that we can use it more confidently as well. Perfect. All right. I better wrap up here because we're running pretty long. It's been a great episode and loads of great insights. Lara, Robbie, thank you so much for your inputs. It's been really, really interesting. These items will be going out as a series over the next week on LinkedIn.

[00:37:40] So if you want to have your say on this and give your reply, then hit us in the comments there. Well, than that, thank you so much for listening and for tuning in. We'll see you again next week. But don't forget to nominate your candidates for the Digital Disruptor Awards later on in the year. The nominations are open. You'll be able to find the link in the episode description. Thanks for joining us again. And we'll see you again soon. Thank you.