Jonathan Sanders, CEO at Light: Light replaces fragmented finance tools with one platform

0:00Jonathan Sanders:Hello and welcome to the Scaling Europe show presented by Deal. This episode is sponsored by SurrealDB, the multimodal database for AI agents. Omni, the AI analytics platform trusted by growing companies like Perplexity and Symbesia and VentureComet, the platform that gives startups and scale-ups real-time equity tracking, daily business insights and automated management information. Thank you for joining me. Please like, comment and subscribe.

0:33Jonathan Sanders:Hello and welcome back to the Scaling Europe show presented by Deal. I'm Seb Johnson. Today I've got Jonathan from Light. Light is an amazing AI-first company disrupting the world of ERP and traditional finance systems. So Jonathan, thank you for joining me today. How are you doing? Thank you for having me, Seb. Wonderful to be here. Look, for those who don't know you or what you're built, can you give a quick introduction? Yes. So Light is an AI-native finance platform designed for allowing companies to scale globally. We provide an end-to-end solution that allows you to run payments, both in and outgoing, accounting, bookkeeping across 100 to 100 different entities.

1:10The consolidation that comes on top of that, including all of the regulatory reporting, such as annual income statement, balance sheet, EAT returns, e-invoices, et cetera. We service a wide range of primarily technology and service companies across Europe and the US, including some of the fast-growing, high-scaling companies like Lovable and Nagora. Amazing. And who are you displacing? We typically either come in and displace a patchwork of four to five different sort of accounting systems. That could be your Zeros, your QuickBooks, your Fort Knox, your Dativ, etc. and maybe one or two or three more point solutions in either the AP, AR, or spend management space.

1:58So we typically see a reduction in number of systems go from, call it, three to eight to one. And the other sort of side that we replace is when people have already sort of established an ERP and they are going from like a NetSuite or a Sage or Workday or an SAP into Light.

2:18Jonathan Sanders:And how different is your product experience, I guess, compared to one of those traditional ERPs? So the traditional ERPs, we wrote like a lengthy manifesto about it, are predominantly created from this world of, okay, I'm going to own your entire business from CRM to HR to finance, etc. And we take in the view that, okay, we'll let the CRM providers do CRM and we'll let HR do HR and focus only on finance. And on doing that, you take a look at what actually goes on of our activity in the finance bucket, and you see, okay, like most of the actual user-facing flows is in spend management and AP flows, and then there's quite a bit in the AR space as well.

3:04So why do you need a lot of these separate different solutions for that? And so instead of having like a broad solution, we have like a deep vertical solution around the finance space where they can get like an accounting ledger, subscription management uh a spend management and a full bill payment that works for you globally and so that's uh in a way how all light is very different now with the age of ai that is becoming a much more powerful proposition because what you have is basically you have a bunch of different very powerful tools that you can construct different agents around and have them operate on

3:37Jonathan Sanders:amazing i know it's clearly resonating with customers like you said you mentioned you've mentioned some amazing names of customers you've acquired you also raised a large i think it was 13 million dollars series a in september last year led by ballston so it was obviously resonated with investors as well i want to touch on that round like how how did that round come about and i guess what did those what what gave those investors conviction to to back you and the team yeah so i think a big part of it was that we had this fundamental thesis that the sort of finance stack was way too defragmented.

4:18And if we could build a very powerful platform that allow you to do these different operations that would resonate with the market. And also that was, you know, the product that I wish that we could have ourselves in prior situations instead of having these, like, you know, you have this NetSuite and then you have like four or five different point solutions around it. And you always have like issues in getting the data to work across the different solutions. So we laid out that thesis quite early in the company's journey and we've been building towards it since day one. And I think, you know, over time, as we catch up with different investors, they saw that thesis sort of being proven out to be true.

4:58And so we've ended up not only the Series A, but also before with like quite strong non-interest before we were even gearing up to sort of start like a process. And so that's basically it. And then we have the fortune of knowing the Baldurton team for many, many years.

5:17Jonathan Sanders:Amazing. And what kind of investors are they? They get partners? Yes, they're brilliant partners. They really see and share my vision or my core belief that this is probably the most fundamental shift that we've ever seen in the history of mankind. I think it's basically we get fire, printing press, and artificial intelligence. And those will be the three big things when you look back on the tech tree of humanity. And so it is wonderful. I wasn't around for fire or printing press, but I am around now for the rise of intelligence. And it is wonderful to actually work with that and like, okay, what is actually possible now that you have infinite intelligence 24-7 that can just work for you?

6:10Like, how do you harness that energy? And harnessing that energy is one of the most fundamental questions in, I think, humanity, but certainly also in software. And so they share that fundamental excitement and joy. So for us, that means, okay, how do we think about building the product? But also means like, how do we think about like, you know, setting ourself up organizationally, you know make different functions will come together uh all functions will split apart people will need need need to know new ways of doing things uh um everybody needs to understand you know if you want to lead uh you need to really understand this technology and really lean into this technology so you can get your hands dirty and have a intuition of like okay what is actually possible now so you can then like bring your function forward um so for us it is a way of operating and it's a way of building and it's a way of building towards that that central vision we've had now since the beginning of can we give you something that is truly organic and i.e can we give you like a software experience where you said you don't need to tell it what to do you just interact with it and it'll from that interaction understand who you are as a person and how you want it to behave and mold itself towards you and so that's what we're building just with the complexity of we're building it for companies which is like a amalgamation of many different individuals and many different teams and and building that experience of creating something that learns from them mold itself towards their needs and and how is that going to change i guess

7:48Jonathan Sanders:now you raised a big money at the end of last year you're looking forward to 2026 and beyond what what have you got kind of coming up in the product you know what are you excited about building and changing and deploying and evolving as you know you've got this bucket of money you're growing the team yes yeah what are you going to be doing with it so one is like we we extended the the the the tools of functionality of the platform so we added more e-invoice networks we added more tax engines we added more countries where you can get out of the box compliance and and and regulatory reporting so that's like one dimension and then the other dimension that we're really focusing on on this year is the entire mission of getting to organic software which is basically for us three different things that is needed.

8:33It is memory, and it is adaptability, and it is self-referenciability. So memory is the fact that the platform can store knowledge about you as an individual, about you as a team, i.e. let's say that you're in the marketing team, and about you as a company. So there are certain things that is also true about you as a company. There are certain things that are true, but it's only about the marketing team. And there's certain things that are true about you, but it's actually only relevant for Seb. And we need to have those memories at three different levels. So we can always give you the right context.

9:09And by you, I mean the software or you as the user. And then it's adaptability, which is a much more difficult problem of, okay, like how do we automatically update memory and automatically do suggestions for Seb based on his behavior? Like Seb doesn't need to tell us what he wants to be true. we should automatically infer that from his clicking around and doing certain actions in the platform. And that's still a motion in progress, but we have some really interesting prototypes and things we're doing internally that we're going to roll out very soon to our clients. And then lastly, it's about how can the system look at this entire setup and understand what's going on and how do you build the rights of a guardraiser on it?

9:57This all sounds a little bit maybe abstract, but it is fundamentally, since day one, we have been thinking about how can we actually build what would be, when we started, a student assistant that works for you. And I think now it's more like a senior associate that sort of works for you 24-7 and learns from you. And how does that senior associate in the form of a product actually work? And I think you can see that also in the public stock markets that the whole software dynamic is undergoing a shift. People don't know where it's going. So it is a little bit of fear creeping into traditional software valuations, which is completely understandable.

10:42But I think it's an indication that the future is extremely interesting that you have this sort of high volatility instead of what I lament a little bit before we had maybe from 2014 to 2021 a long period where it was like SAS was the era of playbooks it was like this is how you built this and it was almost down to the UI level like okay if you want to build that pattern or that flow well this is how it needs to look like and this is how you do it and if you want to like you know grow in that area this is these are like profiles you hire and everything and it was very like playbookish and now all of that has been sort of well if you really want to harness this new technology you got to throw all of that out the window and reinvent it all and so there's a lot of learning in this phase there's not of like experimentation there'll also be a lot of failure but i think this is the most interesting time to be building since i ever joined the world of of

11:43Jonathan Sanders:startups and how do you stay on top of the changes you know the world of ai is changing so quickly whether it's like the models themselves the capabilities of genetic technology is like completely gone off the rails in the last six months how do you as like a founder building you know relatively early stage company in ai stay on top of all of these changes i think the most important part is to have a really strong inspiration about the technology so that you can look at the different parts of your business and say here i don't think we're doing enough. I think here we're doing the right things.

12:16And I think here we're on the wrong track. And I think the only way you do that is to really get your hands dirty and really work with the technology. And it's only by working with the technology and doing stuff that matters that you get a sense for, okay, is it good? Is it bad? Where is it good? Where is it bad? How do I actually do it. And for us in software, the stuff that matters is what happens in production. So Saturdays and Sundays, I am spending a lot of hours working with predominantly Claude and thinking about, okay, how can you use this to actually improve either certain things that I'm annoyed about or do certain things that I wish was there in the product?

13:05Okay, how can I create a prototype that does that how can i really stress test it and the amount of things that you as a founder can do like most founders i think start with some sort of product vision of like oh i wish this existed in the market because then you know this and this would be possible and it would be 10x easier and that's certainly the way that i started light um and so saturdays and sundays when i have a bit of time i'm spending an awful amount of hours and credits on on building basically prototypes or even bug fixes that make it all the way into production. And I think that is one way of actually understanding, okay, how much is possible with this technology.

13:47When ChatGPT first came out, I took, I guess, a course on AI just to understand, like, okay, how do we actually build LLMs? What is it about? In my prior life, I was a quantitative researcher, so I spend a lot of time in the old sort of machine learning world with K-means clusters and all of that sort of old stuff and then you spend time learning about LLMs and you see, oh, it's very familiar to the old world it's a little bit further away from stats and regressions but it's not that far away from K-means we just have the sort of loop this sort of like yin-yang feedback loop where we're like wrapping itself.

14:33And that was one of the most interesting things about using the new technology was that it is very, very similar to some old technology. But what was possible was the amount of data you can crush food to basically get better and better parameters. And if you got like the right exact parameters, there's like almost infinite possibilities with what you can actually get from value from the output. So for founders looking into it, or I would say like for any leader of any functions, you really need to understand this technology. You can only understand it if you spend a lot of time on it. You cannot delegate it because I think it's really hard for you to successfully transform your business or your function from the old ways of working to the new way of working if you don't have a really, really strong intuition for what is actually possible here.

15:31And I think 80%, 90 % of the value from this new technology will come from us doing things that we didn't do before, that we didn't think about doing before, but that is now possible because the intelligence is cheaper, faster, and 24-7 available. It is very easy to get like a bucket list of like, okay, I do these five things, and I should automate these five things using AI. And that is certainly valuable. But I think the true value could be coming from doing things that you didn't even think about doing before because now they are possible.

16:04Jonathan Sanders:Yeah, interesting. You just got to stay so close to the technology as it comes out. This idea that you can sort of be a senior leader and have your team be close to these new advancements is just no longer true. You've got to be on the forefront and the front line of these releases. You've got to be trying and testing all of these tools which is very different. You really got to because I think a lot of the unlock will also come in a company when you change your ways of working. It is so difficult for anyone but the leader of a function to say, now we need to change our way of working. Like Seb, you're not responsible for X, Y, and Z.

16:40You're responsible for Y, and Z, and B, and A now. And this is how you should do it. And we're going to remove this step that we had before. We're going to create this new function that we didn't have. And this new function is responsible for A, B, and C. Like whatever that looks like for you. But I think it's really hard for, you're just outsourcing it and say like, okay, I got this like IT or AI specialist and he's like just gonna like, the only thing that he can do is go through your existing problems and figure out his way of solving it. And there is some value there, but the true unlock comes when it's, okay, we used to be organized as, you know, like this and this and this, and we used to do this, this, this, this, this.

17:19but to really harness this energy of that technology we need to do it like this and this and this instead and that's a totally different way of of working yeah we're seeing that internally we're seeing like wait but what does it actually mean to be an engineer now that you know claude writes like 90 of the code uh uh uh what does that future uh look like and and and and and was For some function, it's exciting. For design, they're getting new tools and powers to express themselves that they didn't have before. It's the same for PMs and for founders. But for others, you're creeping into their territory.

17:58And so you need to have a really strong view of a leader of what does the future look like and how are we as an organization going to get there?

18:05Jonathan Sanders:Yeah, it's really interesting. You mentioned that 90 % of your code is written by Claude. Every founder I speak to, the number is always over 90%, is now written by AI. And yeah, I mean, it's crazy. I mean, even six months ago, nine months ago, that wasn't true. People were playing with it and they're using it a lot. But now, you know, I spoke to one founder who said that so much of the engineering time is now spent on prompting. Can they write an amazing prompt? How do you structure and write this amazing prompt to get the most out of the LLMs and the models themselves as opposed to writing the code?

18:38And these jobs are just changing rapidly.

18:43Jonathan Sanders:I want to touch on growth. You know, when you raised her around, I saw that there was like 30x growth. Where's your growth coming from at the moment? I want to know specifically regions. You know, I know that you're focusing a lot on America. Where's your growth coming from America versus Europe? And then secondly, what type of client and how is that changing? Have you moved from smaller, scrappier, younger startups to larger companies and you're seeing a change or transition in your ICP? Yeah, so most of our growth in terms of regions, we get still the majority of our growth from Europe. And then a healthy portion comes from the US.

19:19In terms of industries, verticals, we are moving upmarket. We are seeing more complex companies moving over that are maybe working from like a workday or SAP setup coming into light. And it's a longer technology and implementation sort of process where we need to work with multiple different sort of project managers, external IT consultants, et cetera. So we are going through that change. In terms of the differences between the regions, one thing that I thought before that is not intuitive is I thought the US would be higher ACV because they had more money, higher GDP, basically. We're seeing the U.S.

20:08What I think would also hold true before was that they move faster and we're still seeing them move faster. But when it comes to commercial pressure, we're seeing U.S. buyers being much more price sensitive than European buyers. And we're actually looking into a market that in the US alone might be slightly overfunded over there. So we're seeing some competition in the US getting a lot of funding. And it seems that the majority of that funding has gone to undercut each other. And that's certainly a dynamic that I wouldn't have expected before going into the US is the amount of price sensitivity and commercial presser.

21:02Not only on finance techs, but when you speak to CFOs over there, because they're the budget owners, it's across all of the different SaaS applications that they use. It's a market that is becoming extremely commercially minded and price minded.

21:21Jonathan Sanders:Yeah, I guess it's just a more mature market and a much more competitive market. That's interesting. I wonder if that's the way that Europe is going to go soon. Look, we're running out of time. I want to touch on, you know, what are your plans, ambitions, visions for 2026? By the time you get to the end of 2026, what will you need to do to make the year a success? Yeah. I think number one for us has always been to get to our vision of building, like, this is the product vision that we have. This is this organic software that molds itself to you. And we have a series of concrete steps that the team is working on in terms of to get there.

22:03And if we get there, then, you know, I am very, very, very, very happy. I think the number one danger for us in this new shift is that we build, quote unquote, the design solution known as Sketch and somebody else builds Figma. And what Figma got right is, of course, in the cloud era, you can do multiplayer and has all of these benefits, et cetera. And it took the competition like too long to basically catch up to that. And if you ask somebody like they never actually had, and that is because the way that you design your product to work, to be delivered, to be serviced will dictate your go-to-market motion.

22:49And if you try to like re-ramp all of that, if you have a really, really, really, really, really big customer base and a very, very, very established, you know, partner network, et cetera, that becomes much harder. And I've seen that in countless prior companies where we had this basically phenomenon of portfolio debt where we knew that if we did something to the product, it would be a better experience for everybody, but it would entail like a large change management with a lot of existing clients. And then we always ended up not doing it because it would be too much of like a quote-unquote risk towards the back book of clients you have, right?

23:25So what we're really nailing is, or what we're really focusing on is like, how do we really nail the product vision? And how do we, at the same time, get our entire go-to-market machine ready to sort of really scale that product vision? So we're looking forward to a year of building out the vision of organic software, which I think will be a household name across not just finance software, but all enterprise software business will need to have this phenomenon because I think it's a better product experience. And how can we combine that with scaling up our commercial focus across the US and Europe?

24:05Jonathan Sanders:Amazing. Well, look, it's going to be a big year for you guys. So looking forward to seeing how you go. And keep me informed of any updates or changes or news. I'd love to help provide visibility and provide coverage. But look, Jonathan, thank you so much for joining me. It's been great fun to chat. Thanks so much for having me on. Thank you.