But What Has AI *Actually* Done For Us? | AI Adoption & Accountant News This Week

[00:00:00] This Is Going To Be A Very Topical And Also A Very Tricky Area For Accountants To Overcome. Please, please, please don't just take what has been drafted and throw it into your social media account. What has been found out is that their AI isn't AI at all, but it was a whole bunch of people in India, 700 people in India who were basically doing the coding in the background. How the hell do I keep on top of this absolute rocket ship of a technology, which is changing so quickly.

[00:00:24] Your team was operating fine when it was 10 people, but then you get to 20, 30, 40, 50 employees plus, your business gets more complex. People are spending a lot of time exporting information, building a quote in a spreadsheet, trying to resource a project in another tool, tracking time somewhere else, exporting the timesheet reports, reconciling it in spreadsheets. It just gets a bit chaotic and it's really out of date.

[00:00:48] This is a product that has been trained on 10 billion pounds worth of transaction data that they've managed to accumulate. If you're not using AI to some extent for marketing, you're probably behind the curve now. Hello and welcome to the Digitals in Accrual World podcast, a place to go for all your insights on the UK accounting news.

[00:01:10] This episode is brought to you by Xledger, the mid-market finance tool that is empowering those accounting firms that want to bring business process outsourcing to their larger clients. We are also going to be diving into what makes Scorotik as a professional services management platform. So that'll be something we dive into a bit later with a key member of the team from there. But before we do any of that, I want to welcome back John Toon to the pod.

[00:01:40] John, we've missed you. How have you been, mate? Yeah, very well, thank you. Yeah, I feel like I've missed loads of episodes. I don't actually know how many I've missed now because I've lost track of where things are at. But yeah, it's been a busy, busy few weeks. And the most recent trip was over in Amsterdam with the data snipper guys for data snipper connect, which was very exciting. I mean, I know it's an audit tool, so maybe it doesn't necessarily appeal to many people in the accounting space. But yeah, they're doing some incredible things over there.

[00:02:08] And yeah, I've got some really interesting releases that are coming out soon. And I know that they are trying to spread their wings into other areas of accounting and finance aside from just audit. So yeah, I'm sure you'll see more and more of them over the next few years at events. But yeah, that was good. And Amsterdam was very nice and it didn't rain until I left. So that was very good. And did you peruse the streets of Amsterdam? Peruse or cruise was that? Whichever, John.

[00:02:38] I did peruse the streets. I did perambulate around the streets of Amsterdam. Yeah, I didn't take part in any of the extracurricular activities that most people will take part in in Amsterdam, if that's what you're trying to allude to. So yeah, we had quite a late night out being entertained by the data snippet guys on Monday. So that was good enough for me.

[00:03:05] So you enjoyed beer with a lot of head, but you didn't go and see the sex museum, which is obviously what everyone goes to look at when they go to Amsterdam. Correct. Although I was in Prague for an event last year and there was a sex museum there. So maybe this is a European thing. Maybe this is what London should aspire to. There should be a sex museum in London. Or maybe there is one. I just don't know about it. I'm sure we're way too prude to do that.

[00:03:33] And I do feel you top trumped me in English language earlier when I used peruse and you came up with something perambulate or something which I've never heard of. Oh, well, there you go. You've won up to me. Yeah. But unfortunately, it appears you've replaced Indy because we were without Indy for this episode. And I'm starting to wonder if you two are the same person. I mean, having met you both, I shouldn't think that, but we just can't seem to get you in the same room at the moment. I know. I know.

[00:04:00] It's definitely like we've gone back to this kind of like one in one out that we had at the beginning of the year when I was so busy and I was on my travels again. It's outrageous, isn't it? But yeah, I think poor Indy is quite busy at the moment. So she's got a lot of things on her plate. She's keeping her entertained. So I think she's got a genuine excuse not to turn up on the pod. Unlike most of the time when I don't turn up, there's probably not really a genuine excuse. Well, you're doing research. That's what we really say. You're going out to these various geographical areas and doing research.

[00:04:31] And I'm sure at some point we'll see an expense claim for it. But it sounds like you've got some insights into Data Sniffer that you probably can't share at this point. And as soon as we can, obviously, we'll put them out on the pod. Yes, for sure. I mean, yeah, I mean, they've got some stuff that's coming out in July, which is a new product release, which I think they'll start talking about fairly soon. And then on top of that, in the not too distant future, they've got some other exciting, fun stuff.

[00:04:57] And I would say one of the things that they are releasing, because I genuinely don't know if I can talk about it. One of the things that they are going to be releasing on the roadmap, which hopefully will arrive around about September time, is probably one of the biggest pain points that accountants and auditors have to face, particularly when they're dealing with more technical sets of accounts, I would say. So I'll leave that dangling there for the time being. Very, very interesting. Can't wait for us to hear more about that.

[00:05:22] But that does sound like the perfect segue into app news this week. And I'm going to cut, I kick this off with, I guess, what everyone's talking about at the moment, which is AI and accountancy. And if anyone's been following me on LinkedIn, they would see that I'm sceptical about what you can do.

[00:05:46] But there has been a great article that's been released recently that does dive into some of the ways that accountancy firms are making use of AI and maximizing, I guess, efficiency inside their firm by using very effective tools. And I'll run through some of these. I'm not going to dive into this in too great detail. You would have seen this on the newsletter, and you can dive into the specifics of that if you want to know the detail.

[00:06:10] If you're not subscribed to the newsletter, please take that as an action point from this news article, because some of the times, actually, the detail is what's really important on whether you can apply this tech, these solutions, or not inside your firm. So case one, as they put it in the article, is about enhancing automating your email. And this is something I've heard a few firms doing.

[00:06:35] So essentially, you can build an AI bot that connects directly into your email. It will see what the context is, categorize it into a folder on what's critical and what's not, and also draft a response based on the nature of what's been received. Some of that will be very quick responses, i.e. it's just like, as you probably get lots of emails, just say noted.

[00:06:57] It may put it away to read later as a research document, or it may draft a quick response based on technical information it can gather from the internet. Some of these tools are using Power Automate to do it. You've got others that have got some very specific, dedicated email tools out there that can also plug into your mailbox and do this for you. But generally, it's all using large language models. And this is something that I believe is becoming more prevalent out there.

[00:07:25] And so I think will, at some point, be kind of the norm inside all email providers. You've then got one that's quite advanced, which is about drafting technical accounting memos. This is where a firm that, I guess, would provide advice to their clients as a memo style would have been creating these for years. They took all of these, threw this into a large language model, so it ingested them all.

[00:07:52] And then every time a new one, a new query came in, it would be fed directly into that model, and it would draft a response based on the information you'd learn. Now, the change, I think, for a lot of firms is that they may have these memos, but they're probably just in a complete mess of a document management system. So trying to extract that information out and put it into a large language model will be quite challenging. This is where I think AI agents later down the road, probably back end of this year, start of next year, will come into their own.

[00:08:17] Because as long as you've got your data in a system in a relatively structured format, then they'll be able to do something with it. If you're not at that point, you need to be starting to move your documents into a structured format, preferably in the cloud, to get the most from them. Another one standard one, there's something that I think probably every firm is doing to some extent, is marketing. If you're not using AI to some extent for marketing, you're probably behind the curve now.

[00:08:43] If it's not on generating images, there's a minimum, or you can just create some prompts for ideas for marketing articles by just throwing them into the likes of ChatGVT. That is something that you'll see a lot of people are doing. The one caveat I would say to this is please, please, please don't just take what has been drafted and throw it into your social media account. It's very obvious now on what is a, I guess it's called a ChatGVT article or post compared to something you create yourself.

[00:09:12] Use it to create or trigger ideas, even potentially draft it. But please edit it to make it feel like it's come from you or at least a person. Fourth, data analysis with AI-assisted coding. Excel. How many people are doing more advanced stuff in Excel nowadays? And how many of your team feel that they can't do this? So you want to pass something down to lower members of your team, less experienced,

[00:09:38] but actually their Excel skill set prevents them from doing this. You can generally ask any AI agent. Obviously, ChatGVT is the one that normally comes up in this, but any AI agent to go, this is what I'm trying to resolve. Give me the Excel code to do it. And it will give you the exact Excel code that you need. Google does it to some extent now because they've built AI into it, but actually going direct to a ChatGVT or alternative model will create you a bit of code that sits in Excel that can solve your problem.

[00:10:08] This is a big game changer, in my opinion, for those that work quite heavily in Excel still and do a lot of formulas in that, whether you're in industry or in practice. The fifth option or fifth area that is flagged in this article is brainstorming and research. Brilliant thing to just come up with either ideas or also to bring back some general information on what you're trying to look into. So many times we historically went through those big, thick books of tax legislation.

[00:10:36] Generally now they just sit propping up your monitors now on your desk and you've been kind of going to the internet for a long time. Actually, the most modern way to do this is using large language models. Feed in what you need to do. It'll generally bring back the answer, but the best thing it does nowadays, it brings back where it's got that answer from. So you can dive into it and confirm or check that what it has brought back is realistic. Sometimes it's going to come back with ghost responses, ones that are not correct.

[00:11:05] It's hallucinated. You need to validate that based on your experience. So good once again for prompting and coming back with initial research answers, but please do not take that as fact. You need to be professionally sceptic as we are as accountants. And then finally, something that plays into what John was talking about in our introduction, audit. Audit automation with AI is critical. Searching for documentation, extracting that information,

[00:11:32] and also creating or covering some initial assessment of those documents on whether they're accurate or not is something that's being built into a lot of audit tools and something that John could probably talk a lot more about than me. Yeah, thanks for that, Ryan. And I mean, there's a great summary. And I think that there's actually so much more that you could do as well, which we could probably talk about this for hours and hours and hours and just list off a whole bunch of things. But yeah, I think as a baseline starting point, fundamentally,

[00:12:02] those kind of use cases that you've talked about are the bare minimum, really, aren't they? Nowadays, I think it's what everyone should be expected to do. Everyone loves it. What's it called? An M-dash? That's the giveaway for a chat GPT generated LinkedIn post. I mean, who knew that dashes came in so many different variations of varieties? Because I certainly didn't. I did not either. But I disagree with you, John. I don't think this is the base level. I think these are examples of where firms are doing quite advanced stuff.

[00:12:32] I think most firms are not doing any of this. The most advanced bit I see a lot of firms doing is they've got an AI note taker that's plugged into their meetings. And you need to start somewhere, in my opinion. You need to be interacting with AI, but don't feel you need to do everything. Focus on a pain point that's really prevalent to you and try and resolve that. And that will kind of prove to you that AI is really effective. But depending on the size of the firm that you have, it really depends on how much you can invest in it. I know that obviously your team,

[00:13:00] you have created a team of being in-shubbers that can focus on AI, but you are at a reasonable scale that allows you to do that. Much smaller firms can't do that. And they kind of need more out-of-the-box answers. And I feel we're not quite there. Everything is very bespoke at the moment. Until we get to more generic solutions, I think actually the majority of accounting firms are still kind of trying to find the way forward. Yeah. I know. Yeah, I wouldn't disagree with that.

[00:13:26] I mean, I think the one thing that I guess we have seen in accounting bookkeeping space for a long time, right, is that we've had vendors who have come to us and said, here's a solution that we've built. You run away with it and take it. And if we're being sort of slightly skeptical, we know that some of those solutions have not had a problem that they're trying to solve. They're trying to create a problem to solve with their product. But the ones that have been most successful

[00:13:53] have been the ones that have sort of sped up the time it takes to do the bookkeeping, you know, an OCR and extracting invoices or time to pay and credit control and all this fun stuff that we're in the middle of. And as we've talked about many times, we could wax lyrical about AI and accounting. But there's other stuff we need to cover. Indeed, indeed. So let's pick up a bit more news then in terms of where things are at.

[00:14:20] So something that totally, totally blew me away was that Airwallex have raised a 300 million Series F to continue investment in their platform, in their product, which takes their valuation to $6.2 billion. I mean, I was blown away by that. Maybe I've just not been keeping up enough with these things.

[00:14:47] But I was like, wow, $6.2 billion is a huge valuation for a platform that I still think of as quite new and everything else. The round included some additional secondary share transfers between some investors and some of the shareholders, which counted for about another $150 million. And again, another big figure is that their total investment so far has now totaled $1.2 billion.

[00:15:14] So Airwallex are doing really well. And they've got a whole bunch of big and famous faces in here. But I guess the one that really stands out is probably Visa. So Visa Ventures have joined as a strategic investor. And I guess they'll be then looking to help them to kind of push their platform out even more globally than they already are and reach into more markets and stuff. So yeah, I mean, first of all, congratulations to the Airwallex guys.

[00:15:42] This is absolutely staggering and what a great valuation. And clearly there's a huge appetite for people looking to continue to invest in their products and their platform to scale and grow. And isn't it nice to see and hear about a product that isn't using AI necessarily or isn't promoting their use of AI in their platform still receiving investments? So it just shows that there are other technologies and platforms available that you can invest in and still get a good return on. Completely agree on that.

[00:16:09] My understanding is that the reason they've taken on such a big investment is to try and build more into actually having a full banking product. So they've always been kind of that intermediary that sits between bank accounts. You can obviously still use them as bank accounts, but you wanted to kind of own bank account from a security perspective that sat alongside it. And I believe they want full banking licenses and to build out the functionality from there. So they're really going big on it and be interesting to see

[00:16:37] if, yeah, what exists, the new kind of clear bank over time. Yeah. We'll see. I've got something on a very different area. Caseware have acquired Lease Java. Now, I'd never heard of Lease Java before, but essentially it's a way to manage your leases inside a cloud-based system. And the reason they've acquired this is they're going to build Lease Java directly into the Caseware platform.

[00:17:04] And using the AI, Caseware's AI-powered digital assistant, ADA, you'll be able to extract key terms from those leases for reference inside the accounting and audit product. So with leasing becoming, well, still quite a complex area with lots of changes, I remember from a financial reporting perspective, leases were always very complex. Having, I guess, a lease management tool directly inside the, or, you know, as part of the Caseware suite,

[00:17:34] I think adds a lot of value here. Now, the nuance, you can, you know, you could say, oh, we can have any document management system that works with leases. Lease Java is specifically built to handle complex and nuanced scenarios in leases, including partial termination on weighted average calculations. So if you just got standard basic rental leases, there's probably not something that's going to add a lot of value. But if you've got very complex lease arrangements for your accounting and audit clients,

[00:18:02] actually, this is something that would add a lot of value covering from a financial reporting perspective, mostly the IFRS 16 requirement. So I think this is something that's mostly gone under the radar. I've not seen a lot of people talk about it, but could add a lot of weight to a case where, especially around that audit side from the scale of clients, their complex lease arrangements. Yeah, I mean, I think this is a really great move from Caseware, because particularly with the FRS 102 changes,

[00:18:29] which are going to bring in effectively IFRS 15 and IFRS 16. So that's the revenue recognition changes and the lease accounting changes that you mentioned. This is going to be a very topical and also a very tricky area for accountants to overcome. And just having this baked into your kind of accounts production and your audit products is a bit of a no-brainer. We've seen a few other people in the market doing similar things. They're still independent, but maybe they'll get snapped up by someone else.

[00:18:59] And this is a problem that's almost like twofold insofar as it's going to be complicated and challenging for us as accountants and auditors to kind of overcome. And probably bookkeepers as well are going to be involved in some of this. But equally for our clients, particularly those that run their own finance or at least try and manage some of their own finance, they're going to want to be able to have access to some of these things. So what would be interesting here from Caseware's point of view is whether they intend to make this available to commercial organizations rather than just accountants through the platform.

[00:19:28] And if not, I guess there are other tools that have taken that view that they will go direct to market or they will go through a partnership arrangement with an accountant. So yeah, interesting times for sure. I've got something from Iris. We know that Iris are playing big in the US. They've got a huge focus and that was a stated ambition after their most recent PE investment round, which feels like an age ago. And I'm not even sure, have they appointed their new CEO?

[00:19:58] Because I know they were still looking for a CEO to be based out of the US to help sort of build and grow. But maybe we'll pick that up separately and I can refresh my memory on where things are at. But as of just over a week or so ago, Iris have picked up and bought Instead, which is an AI-powered tax platform aimed at helping the automation and preparation of not just tax compliance, but also some advisory tools around that as well.

[00:20:28] I mean, this is fascinating for me insofar as, I mean, this is a US product, so it's not that relevant to the UK. What's really fascinating to me is that Instead's a product that's only been around since June 2023. So it's what? Just coming up to being two years old and already they've managed to get to a point where they've scaled, grown and sold out to one of the big players in the market. This isn't new, particularly in the US market. We saw that Thomson Reuters Digita made an AI-based acquisition,

[00:20:59] I think towards the back end of last year. I think it was like October, November time of a product that had an even shorter life. I think it had only been in market for just over 12 months. I think it was like 14 months old when they bought it. So I think this is indicative of a couple of things. One is that we know there's huge pressure in the US market around your capacity from a people point of view. And so firms are heavily investing

[00:21:29] in anything AI-powered. There's always a caveat over that because I mean, I guess if you're really up to speed, then the news that Builder AI has gone bankrupt, which is not really an accounting-aimed product, but it was a product that Microsoft and others had invested massively in, had a valuation of 1.2 billion. And that's gone bust. And actually what it's been found out is that their AI isn't AI at all, but it was a whole bunch of people in India, 700 people in India who were basically doing the coding in the background

[00:21:58] to create the stuff that was going through the Builder AI platform. And so I think we do have to be a little bit cautious about some of these things. So yeah, interesting times for Iris and for Instead. Yeah, and I still can't believe you get more and more of these businesses that are just shells. You know, they're, oh, we're AI, but we're not. It's quite depressing really. But the one thing I'm nervous about on this is that I've seen a few businesses

[00:22:28] try and automate the collection of personal tax information, mostly focused in the UK. And I find the issue is not that it can't ingest it and it can't extract it, but it's picking up what's missing. And it's that kind of skepticism that is built in, you know, professional skepticism that's brought into accounts of tax professionals to go. It feels like something, you know, from, I guess, our interactions, from what you've said in the past, I feel like there's something that should be here.

[00:22:57] And I've not come across any tools that have any clue on how to deal with that. They just rely on what's provided by the individual. And so once we get a good example of that, and this may be it, then, yeah, I think we have solved a huge pay point. But I've got something else that is also US related. So Join, the multi-entity consolidation reporting tool, has integrated into Puzzle. Now, what I thought was interesting about this

[00:23:26] is I've never heard of Puzzle, but I'm quite well versed with Join. And from my digging into it, from what I can tell, Puzzle is a US-based accounting solution. It's really pitching itself against QuickBooks Online, not one that's tried or is over in the UK or has pushed itself over in the UK. So I can tell you from this, Join are really trying to push themselves into being a US reporting tool as well as UK. That's maybe a big plus in the differentiator for them here by integrating into Puzzle.

[00:23:56] So what am I talking about on this? It's just whenever I see an integration, obviously there has been a strategic play to want to do a deeper relationship with another party. And if they're focusing on the US entity, how does that play out for the UK side? Hopefully this is just because they are built up to a new, of a certain scale now that actually it warrants that geographical spread and we won't see anything from any negative impact in the UK. But the integration itself is very standard,

[00:24:25] is looking at saving hours for automated consolidation, which is exactly what Join focuses on. And something that, similar to I guess the likes of Quibbux Online, Zero, etc. in the UK, which is very siloed on single entities, Join, I assume Puzzle is in exactly the same way and Join will resolve that challenge by wrapping around multiple entities and sucking all that data in to do that multi-entity reporting consolidation. Yeah, it's an interesting one.

[00:24:53] I'm going to pivot away from that though and move into a fresh brand new product out into the market and one that's got some very high ambitions it would seem. So there is a brand new product in the market called Ravicle, which launched a couple of weeks ago. I know you're going to talk about some of the investment and stuff that they've managed to pick up, Brian. But the background to this is that this is three guys that have spilled out of Silverfin.

[00:25:22] So Joris van der Goet, who was one of the co-founders of Silverfin and exited, oh gosh, what now, 18 months or so ago, I think it was, when he left approximately. And that was off the back of the investments that they'd had and then ultimately selling out to Visma. And he's joined by two other people that were part of the Silverfin family, Ben van der Meere and Ken Bastiensen. Both of whom actually ironically came into Silverfin

[00:25:50] from a previous acquisition that they made of an AI business, which is more focused on machine learning than LLMs. So a different part of the AI umbrella, I guess. But Ken in particular, I knew reasonably well and he was sort of, or him and the team were the powerhouse between Silverfin's chart of account mapping and that automated functionality that they've got in the product there. And so, yeah, this is really exciting. As it happens, there was some coverage in the Belgian press,

[00:26:20] which unfortunately I can't translate because my Belgian is pretty, pretty, you know, terrible. Well, not terrible. It's non-existent. Let's put it that way. But basically, they've been they've been touted as Silverfin toppers, you know, they might be able to displace their position in the market, which, well, I know it's bold. It's bold, not just from Yoris, but also from the team there. And the whole premise of what they're building, you're perhaps going to be unsurprised to hear,

[00:26:50] is all around building AI agents that are going to be able to perform a whole bunch of different tasks. And actually, that doesn't look like they're going to be focusing purely on accounting. They're talking about Google, but in HR and payroll as well, insurance, and a couple of other sectors. So, sounds like they're going to be fairly broad-based in their approach, potentially. So, whether that's going to be a strategic advantage or a disadvantage, it's very difficult to say. But essentially, yeah, they are looking

[00:27:20] to create these domain-trained AI agents that will start to deliver a whole bunch of services. Now, I guess in the UK context, at the very least, this isn't necessarily new because we've got people like Genesis and Briefcase in the UK market doing some of this to a degree. And so, I guess that the challenge for Ravicle is going to be how quickly do they expand in their home market in Belgium and some of the other European countries that they'll be familiar with? And then what do they do when they start to come

[00:27:50] into places like the UK, US, which we know has got a big AI thing going on over there as well, and other places. And I guess the key thing for Yoris and the team will be to learn maybe from some of the Silverfin's failings when they expanded out into the UK in particular is that they really tripped over themselves for a period of time and found it quite difficult to get some traction for a period of time. Yeah, and I guess this is an example, John, where we've come up with similar bits of information

[00:28:19] but different articles, right? Because as you said, I've got some information on the Ravicle secured 7.3 million euros as kind of their initial investment round. And what I thought was fascinating on this is that they've done that with only 10 pilot projects in place. So given that they've not, I guess, got long-term secured contracts, they're only kind of in that test phase, securing 7.3 million euros is quite an achievement, and that must be based on what the long-term objective for that system is.

[00:28:49] And I agree with you that they'll be up against others in this space. What I read on this, it'll be interesting to get your take on it, is that they're very much AI agent focused. So we've had lots of conversations about AI agents, the fact that we've not really seen them in the UK market. You've got the likes of Briefcase Genesis doing things on the OCR perspective. And I know there's some that are trying to broaden beyond that and look at kind of combining OCR and accounting into one

[00:29:19] tool. So deal with banking, deal with purchases, and you've got a lot of the whole accounting package wrapped up in AI. And essentially, unless you start doing period and close as well with AI agents, and unless that's what Raric are focusing on, yeah, they've already got others that are probably further ahead that they're going to have to try and catch up with. Is 7.3 million euros going to be good enough for that? I'm not so certain. I mean, that perhaps is the

[00:29:48] 7.3 million dollar question or 7.3 million euro question, Ryan. I mean, I guess with any of these things, you could bury a whole ton of money into these challenges. But, you know, again, if I kind of reflect on some of the stuff that I learned from Data Sniff and Connect, it was that I had a guy there presenting from Microsoft, but really talking about what's going to be happening with Open AI and ChatGPT. And effectively, Microsoft and

[00:30:17] or Open AI have effectively stopped investing in functionality and features in the products that they're releasing. And their entire focus is around agentic AI utilisation and how do you build out these agents and how do you create these agents. So their expectation is once ChatGPT 5.0 is released, which is anticipated at some point in the summer, it's going to be highly unlikely that we'll start to see more and more of these incremental

[00:30:49] versions of the LLMs coming out. And like I said, they're going to be much more focused in on specific, narrow use cases using these agentic properties and stuff. agents. I think you can apply them to a particular process, like we've already discussed, whether it's the bookkeeping stuff of taking stuff out of invoices and putting it into a GL, for example. But the real opportunity here

[00:31:19] is that you can build and build and build on this because you don't just need an agent that can do a very basic task. And I'm trying not to be too condescending there, but taking data out of one thing and putting it into another is basic. But what you could also have is an agent that does a review or a management type role as well. And that is when it starts to get a bit more complex because that's when you need to understand the nuance and the context. And you mentioned this when we were talking about

[00:31:48] the caseware stuff. Yeah, I agree with all of that. And I think that's where when everyone comes to me at the moment and say I should be doing something with AI. I say realistically, if you pause and wait six months, you're going to see a fundamentally different way that you can utilise it. So work out when to go, when to push, make sure you're set up and your systems are ready to do it, the structure's ready, and I think you'll get a much bigger impact by just actually investing your time a little bit later. Most time I go, actually you need to be doing

[00:32:18] something now with AI. I think six months, you're in a much better place. Yeah. I mean, I think the other thing just to add to that as well is, of course it's fine to wait or it's fine to figure out what these use cases are and in some respects from a strategic point of view, waiting for a vendor, whether it's Ravicle or someone else, to come up with a solution that works for you as a business, that's really great and it lowers the risk. the one caveat I would have on that is that we know

[00:32:49] that the task completion capability of some of the early stages of Agentsic AI that we've already seen and again, you don't really see it here in the UK market because it's not available but in the US, ChatGPT have already got task completion into their products and yes, you pay $200 a month for it which might seem quite steep but actually the impact that that could pay back 10 to 100 times maybe even more back in terms of that investment. We know that our capability is doubling every seven months so

[00:33:18] we're here early June basically by early 2026 the capability of what we see today is going to be twice as good by early 2026 and we've been talking about this for quite a while now and we talked about this back in March April time and so effectively we're almost going to track it back there and so the other side of summer we're going to be in a different place again and this is a challenge

[00:33:48] for not just accountants or bookkeepers but everyone that runs a business is like how the hell do I keep on top of this absolute rocket ship of a technology which is changing so quickly it's so difficult to figure out whether I should use ChatGPT or Clawed or Gemini or whatever it is but also then keep on top of all of the other stuff that goes around in the periphery around this and we've talked about technologies like MCP and we've talked about agentic properties and task completion and a whole bunch of other stuff and that in itself

[00:34:17] is impossible to keep on top of and I think that's why certainly in our market we're seeing people in some cases being very cautious and caution is quite often a good thing but also that does have its own inherent limitations in terms of potentially restricting your opportunities yeah and I think you need to do something and you need to make your team aware of doing something and bring at least some

[00:34:47] level of AI in that they can interact with and play around with three four months time of going this is what we're going to do now and change everything you're actually going to cause massive problems in the internal running of your business so start giving some exposure to AI and utilizing AI even if you can't do

[00:35:17] everything now for sure yeah well do we have anything else on AI John is that it we wrapped up for this week we have one last little bit so we had previously which is more about some more tax AI so counting up hopefully a business that a lot of people are familiar with here in the UK they've bounced around and done a few different things over their time in the accounting space but they are launching

[00:35:47] SmartTax this is a product that has been trained on 10 billion pounds worth of transaction data that they've managed to accumulate and the whole premise of what SmartTax AI is going to do for you is that it will help to automatically categorize your transactions it will flag tax risks in those transactions it will help you to keep things like VAT corporation tax or self assessment estimates up to date

[00:36:17] and of course it continues to learn and get more and more and smarter over time so this is available in the platform there is a new tax tab effectively in the platform if you're familiar with using it and it will be available for anyone that's using accounting

[00:37:05] challenges so maybe having AI take away quite a lot of the heavy lifting is going to help to make that transition a little bit more smooth I guess definitely and if any of you tuned into a recent digital digest episode we covered how Intuit are going to be looking to do that as well which will release some features over the back end of 2025 we predict what a great way to wrap up a very AI focused app news this

[00:37:35] week so I'm excited to be joined by Harv Nagra from Scoro today for those that haven't heard of you Harv what is your role at Scoro my role Ryan thanks for having me on is head of brand communications oh cool just so I understand what does that mean well interestingly I was a customer of Scoro I used to be an ops director at a

[00:38:05] consultancy and so I loved the platform so much that I wanted to join the company and at Scoro we kind whether it sales or services or marketing

[00:38:35] but to build the profile of the brand out in the world and on LinkedIn and all that kind of stuff now there's a great summary and yes I've been on the podcast and it was great you're a great host hopefully I can live up to your standard half in this chat but I've been working with Scoro for it feels

[00:39:05] it's probably one of the least sexiest terms out there professional services automation rolls off the tongue doesn't it it's like the clunkiest thing and it's quite rare that we go around calling ourselves professional services firms so it's not very relatable in that sense but it does differ from project management software quite a lot you know when we think of project management tools it's basically a to-do list

[00:39:34] services automation platform does is it brings in more of that kind of full workflow functionality so what I mean by that is you can create budgets and quotes in the platform you can resource your projects schedule resources and see utilization availability for more work and all that kind

[00:40:04] of their purchases so that your project managers can track their project budgets and of course for finance and operations people it lets you see performance across the entire organization or multi entities but also report on the business to see well first of all you know do you have resources to do the work and are you over burning how much profit are you making all that kind of stuff so it's a platform that does the end to end

[00:40:34] of the workflow including invoicing and of course it syncs with third party tools as well Xero or QuickBooks or Sage or your HR platforms or your CRM systems if you're kind of advanced enough to have all of that other tech ecosystem in place so the way I

[00:41:15] critical operation systems are coming more and more in depth and Scuro is placed well in that sweet spot to support professional services firms and I think that's realistic why I always keep coming back to working with you guys is because you do answer so many challenges and I know we could delve into some of the depths and the nuances of what that actually means from PSA and I'm sure we'll come back to that at some point but what I guess I'm really interested first is why did it come about what were you trying

[00:41:45] to resolve here there's always a problem that a tech business wraps around to resolve and what drives Scuro what is it you're trying to help businesses with I think it's the efficiency accuracy and visibility is how I would put it you know there comes to a point in any service business based businesses growth journey where you know when you started out with 10 or 15 people you were using like an array of off-the-shelf tools

[00:42:15] or maybe even best-in-class tools and that kind of works but then you get to the level maybe 20 25 30 40 50 employees plus and every time you kind of hit a bigger milestone your business gets more complex your projects get more complex and your team which was kind of operating fine when it was 10 people working with that array of tools when you get to that larger size it just becomes more and more inefficient right so people are kind of spending a lot of time exporting information

[00:42:45] from like you know building a quote in a spreadsheet and then trying to resource a project in another tool tracking time somewhere else exporting the timesheet reports kind of reconciling it in spreadsheets it just gets a bit chaotic and it's really out of date and you know I think your audience probably works quite a lot with finance people or consulting in your agency leads and that kind of thing and they will end up spending loads way more time on

[00:43:15] reporting than they need to and by the time they're actually looking at their reports they're inaccurate because all of the manual data manipulation and replication that kind of thing so the idea is by having everything in this one place and automated it becomes real time and much more accurate yeah and you've kind of got this thing where you know cloud came to the for and you've got the plethora of tools and everyone started

[00:43:45] trying to find best to breed and then trying to plug them together and now we're turning that on its head and things are starting to consolidate down and we're getting things back into one solution but it balances quite powerfully as well because you've got a good API you can do a there's native app integrations that are available

[00:44:14] off the shelf so they just speak to each other there's an API that developers can take and build anything with and there's tools like Zapier which work really well in the middle so somebody who's not a coder can plug two tools in together so Scoro works really empowering your team by giving them the visibility via

[00:44:44] dashboards to see what they're doing how they're doing their day-to-day operation so that they can tweak what they're doing to become much more efficient and effective inside the business I've found this very powerful with Scoro not something you see in every tool out there so you're empowering not only the back-end admin team finance but also those that are on the ground with

[00:45:15] those disjointed tools people end up either using them developing their own way to get their work done whereas when you have it central and in a central place and consistent it means everyone's able to use it for what they need and the information is in line with everyone else really right so the projects that you're working on they're visible in your dashboard if you're one of doers you know if you're a strategist or consultant or whatever you see what you need to see

[00:45:45] and you can access that information rather than just having this kind of siloed tool that just has a list of tasks for you to manage or whatever it might be yeah and another thing that I always found quite relatively unique about Scurro is individual entities and you wrap around all of that to create an interface

[00:46:15] where no matter where you're working or what entity you're working at you can still collaborate which I've always found very powerful especially in the fact that I can still plug into different finance systems and push the data to where it needs to go exactly that you know the business that I used to work at as operations director we were multi entity we had an entity in France and Switzerland in UK so those were all managed in Scurro as you know we could switch between them but each of those entities

[00:46:44] was able to work in their local currencies as well and of course invoice in whatever currency they wanted based on their client needs and we could then centrally report each entity could sync with a different accounting tool for example and again that's the key difference between Scurro and a PM tool they're not designed to track budgets and track budgets and performance and certainly not to report on a business on that kind of financial

[00:47:14] level whereas PSA and Scurro is so a big advantage there and I find a lot of these operational tools focus very much on tasks so you're going to do this you're going

[00:47:48] to

[00:48:20] Practices in place to ensure that we get focused work done are really kind of exemplary. Encouraging us to have do not disturb on all day was kind of shocking to me at first. How are we going to get anything done if we've got no notifications? But that was the whole point is that we shouldn't be constantly distracted by the first email or Slack message that comes in or that ping from BBC News, right? So you silence all that stuff. You plan your day and you work on what you're meant to be working on and check messages in between.

[00:48:49] We, nobody typically needs a response 30 seconds after they send you the message, right? So yeah, really some really interesting practices in place at Scora itself. Yeah, and I like a key thing, right? Is that if you're bringing in a new tool, sometimes it's not just about what the tool can do, but the culture inside the business. So you have to look at it holistically. We're going to dive into a lot more aspects of this over the coming weeks. But thank you so much for jumping on. Give us the initial insight into what Scora is all about. Absolutely. Thank you so much.

[00:49:21] And that brings another DigiTools in a Cool World podcast to a close. Thank you for taking the time to listen. If you want to catch or learn more about what we've talked about in this episode, please make sure you subscribe to our newsletter, which comes out every week. And if you want to actually have your say, you want to ask the questions on the key topics that we cover.

[00:49:42] We do have a weekly DigiTools in a cool podcast. However, what are the specific points of interest, the specific requirements and impacts on the accounting industry from this news article? If you do have any thoughts, please do reach out to us. We always like to hear from you and we'll catch you again next week.