How They Raised $6.3 Million: The Story of a Cybersecurity Startup's Journey

0:00Ben Murray: Greetings everyone. I'm excited to welcome Mahdi and Kim from Donguard. Mahdi is co-founder and CEO and Kim co-founder and CTO. Welcome to the show guys. Thanks for having me. Yeah, great to have you here. First, how about Mahdi? We'll do some quick intros, but Mahdi, tell us a little bit about your background. Mahdi Abdulzarak, I'm the CEO co-founder of Donguard. So I have been actually, I have spent my entire career in security only. So I started hacking when I was 15, and I've been doing all kinds of roles in security for 30 years. Before DownGuard, I was running one of the largest security teams in the Netherlands, 140 people in my team.

0:37Ex-co-founder, 20 years ago, I co-founded a cloud competing startup. I teach in cybersecurity and now fully focused on DownGuard.

0:45Ben Murray: All right. Appreciate that. And Kim, how about a little bit about your background? Yeah, so Ben, I've always been in tech myself. So anything from security engineering to security architecture, CISO roles, CTO roles, but always on the technical side of the house. Yeah, I've worked at cool places like the military where we did things like developing malware, but also at large corporates like Amazon working in security there. And really grew frustrated with the fact that we struggle building secure systems. And that's why I started DogGuard. All right, great. Yeah, so a lot of experience there. So let's talk about DogGuard.

1:21Ben Murray: Tell us what products and services Donguard offers. What we do then is whenever you as an organization built new IT systems, you have to design them first, right? So just like you would with a house, you end up making a blueprint, if you will, where you take all your requirements in and you then start building that system. So in most corporations nowadays, this process is quite broken because the people that are making these drawings have to meet so many different requirements across so many different technologies that what we see is that oftentimes designs are not made properly. And even if the designs are made properly, they're often not executed the right way, right?

2:03Imagine you have a contractor on the construction side, you have a beautiful architectural blueprint of your house, but somebody puts a door in the wrong place. We see this happening in IT systems all the time. So what we do is we help you very quickly make that design and then translate that into a real-life system in a matter of minutes, which currently takes weeks to months.

2:24Ben Murray: Okay. And tell us, how about who's your ICP? So give us a little example of a company that would sign up. And are you targeting any company, certain size companies? But tell us a little bit about that sweet spot for Dawnguard. So we target large and large enterprises. So in essence, any large enterprise that has a cloud estate, because we mainly currently focus on cloud estate, we will support on-prem and OT and what have you. So large enterprises, our sweet spot, we do have a weakness for regulated industries. Because we have an intent factory in our platform through which you can actually select any law and regulation, and then we will translate those automatically into how the system is configured in the design.

3:09but in essence we target large enterprises with a security team yeah okay so large enterprises

3:15Ben Murray: that security team and is that then what's the title what kind of prospect are you going after in these companies who are you trying to find out so it depends we have we have we have a dual approach so we can sell to the cio cto but also to the czo the cio cto they they own mostly the cloud, the dev teams. So we have very clear futures for them that they can use to better manage their cloud and deploy systems, manage cost, et cetera, performance. But the CISO, of course, for the security, we have security engineers, security officers. We actually automate the entire security review process. We have a collaborative campus where these different teams can work on it.

3:55For example, the cloud team can design a system and then the security can do the review on it. It's a collaborative campus, live campus. Okay.

4:03Ben Murray: So when you say large enterprises, are we talking, say, a couple thousand employees and above as your target? Yeah. Yeah. 500 plus. Okay. Our largest customer has 70 ,000 employees. Yeah. Okay. So about 500 and then up and maybe as big as it gets, which would be great. As a sizable cloud estate, then we are good to go because we can immediately deliver value. yeah okay and then we'll talk a little fundraising next but what year did you found the company so we were officially incorporated in july last year and we did our pre-seed round in august actually we had to wait for for for the company to be incorporated properly and then opening a bank account before the money could be transferred yeah well we raised a pre-seed of three million and we just last month closed an additional round of pre-seed raise of$3.3 million.

4:58So in total, we have so far raised$6.3 million.

5:03Ben Murray: Okay,$6.3 million raised to date. And so tell us, so you're about a year in, so still pretty early. And it sounds like you said you had some money already coming in, maybe before you had the bank account. So tell us about that process. Did you have the concept? Did you have PowerPoint slides and then you went out and raised some money? Like, how did this start that the money was coming in right away? Yes. Kim and I are, of course, very experienced people in the cyberspace. And I'm myself also, of course, as ex-co-founder and also an engineer investor. So I'm also an active investor in all kinds of startups, mostly cyberspace treaty, helping many of them.

5:39So we know the game, in a sense. We know the problem space. We know what we want to fix and how to build it. That allows you to go very fast. After meeting Kim, we decided that we are going to start a company based on his PhD research into the topic. I think we spent half an hour on a deck, five slide deck, and with that we raised our first receive money. And before we knew, actually, we were bombarded inbound. We spoke with some all inbound in a matter of four weeks or so. So it was quite intense. And at the end, we chose a VC that really invests based on conviction and a thesis that really fits us well.

6:17Yeah, it makes sense.

6:19Ben Murray: And Kim, from your perspective, is this something where it would just as a CTO, like so much easier just to raise some capital and build the product faster versus bootstrapping this type of product? Would it just take too long to bootstrap the build? Yeah, that's what we thought, right? And that's also why we scale our team pretty quickly. So we're already with 30 people. What we see is that the space is moving so quickly, especially with all these organizations adopting AI. We see that the problem is actually growing and growing much faster than we had maybe anticipated, let's say, three years ago when I started my PhD research.

6:54And that's why we really thought we really need to raise money. We need to build this fast, bring on a top tier team and make sure that we get this help to customers as soon as possible. Okay.

7:06Ben Murray: If I heard that right, it sounds like you're just with your experience and your backgrounds, that really helped as far as just raising that initial money versus if you were a first time founder. Yeah, I think what the most VCs acknowledge that they see a perfect founder market fit. let's say because we know the subfield very well and especially corporate security so that really helps to go fast and then of course the experience really makes you do most things very intuitively and very fast and I think one of the things that also helped us a lot is the chemistry between Kim and me due to our backgrounds and common experiences we have such a chemistry that we don't even need to meet or align things.

7:51We do our own things. We just do it. We know what we want to do and how to do it. I fully trust him. He takes care of technology, everything, product, technology, and I take care of everything else. And that really helps. We'll often see founders discussing a week long about some choice. We don't have that. We just don't even align on it. We just do it because we fully trust each other.

8:15Ben Murray: Yeah, I'm sure a lot of co-founders would love that chemistry that you guys have. And then we hear a lot of advice for first-time founders and fundraising. But guys, between the two of you, any advice for, say, that the second-time founder who had a successful exit, built that resume, now looking to raise or start another company? Is there any fundraising advice for that group when they approach, say, investors with just a concept? I think one thing that I would advise, be yourself. At the end, at the early stage, VCs invest in the person. So just be transparent and just be yourself and then also be transparent about how you like to operate and how you like to collaborate.

9:00In our case, because we are defining a new category, it's for us very important to have a VC that is at least not hindering us in what we are doing and is there to advise us and supports us. That was for us the most important criteria. If you see that we would like to work with for many years, and if you see that trust you to be in the driver's seat and do your thing. And of course there are some practical stuff you should always watch out what kind of terms you sign, you know, lip-pref, stuff like that. But so I would advise to engage a good lawyer for that. Beyond that, the best long-term relationship that you could build is based on being transparent about all the facets of who you are and how you perceive things and how you operate.

9:44I think that solidifies a good foundation for long-term collaboration. Yeah. Yeah.

9:50Ben Murray: Yeah. Makes sense. And Kim, any thoughts there as far as the funding rounds and any lessons learned for founders who may be listening? So for me, it was quite a crazy experience, Ben, because I had no involvement with the venture world before I met Madi. I remember meeting him and we were talking to each other the days after we met. And he told me, we could raise millions in a matter of weeks. And I was like, wow, really? Kind of struck with disbelief. And then we started that process. And indeed, we were able to do this relatively easily. And I think for me, one of the things that I would advise is really look for a co-founder that you work with extremely well.

10:31And especially if you don't have that experience yourself, try to find somebody that does. Because without Madi's experience and previous network would not have been able to raise like this. Yeah.

10:41Ben Murray: Yeah. Really interesting. And then who wants to talk go-to-market motion? In targeting enterprises, large enterprises, I'm sure this is, I always like to, I'm always very curious about the sales cycle, but tell us a little bit about your prospecting right now and how you're reaching these decision makers within the security teams. We are just starting our GTMotion. What we have done from the beginning, we engaged a set of a dozen, with more than And there were 15 design partners. Most of them were multinationals that really believed in the problem that we are fixing and that they also have to deal with.

11:15With a very intensive process and co-collaboration, we developed the platform based on the input of the customers. And we just last week made the system generally available. So we onboarded Adolf GTM just five weeks, six weeks back. And he's now actually building the entire, let's say the sales machine, the GTM machine that is repeatable. But I do like what about what he's doing. He's really using a lot of data to identify the right customers, stakeholders to approach. And yeah, he's building the funnel at this moment. But it's quite interesting that we got dozens of inbound booking requests for demos from very large organizations that exactly understood what we are fixing.

11:55So that's very hopeful. Yeah.

11:57Ben Murray: Interesting. And you hear a lot of early stage companies with going the design partner route. So tell us a little bit more about that. And I'm always curious, are they getting a free ride right now with the commitment that they'll sign a paid contract? Are they paying a little pilot money right now? How does the monetary side work with these design partners? So in our case, we chose that route because we are expertise domain. So there was a real risk that we would develop this platform from ivory tower because, of course, we are experts in cybersecurity. So we didn't want to do that. So we wanted to really develop a world-class product, how customers love to use, where it adds value, where it fits the workflows and processes, et cetera.

12:40And we also decided from the beginning that we are not going to ask for money. Of course, in a product that you start building, we had early proof concept. You cannot ask for money for it, but some of our great design partners decided to donate some money. Just a token of support and appreciation. And yeah, so we had weekly meetings with all those design partners. They would go through the bugs, the futures that we released, their feedback, etc., etc. And that was really valuable. And one of the things that was also interesting to notice, that most of the future requests that they were coming up with were already on our roadmap.

13:18It was quite positive. Yeah. And we are now converting those customers into commercial contracts. Some of them negotiated. And as of course, it was talking about appreciation. We gave them discounts on the first commercial years.

13:30Ben Murray: Okay. And I was going to ask, it was interesting because you said you just went GA recently. So is that kind of that threshold? Because how long do you let this go on where it's like, boy, I've got to ask for money at some point. And what is that point in time? So is GA or making the product generally available? well, then is that good time or was the ask previous to that? So it depends per customer. Of course, many of these design partners onboarded at different times. So in essence, they have a one-year design partnership contract where they can use the product for free. And then there was, of course, space to negotiate discounts for the commercial contracts.

14:08And most of them negotiate discounts. So in essence, when the design partnership is due, they do need to confer to commercial contracts. But some of them are already conferting before the end of the commercial. That's great. They're a partnership corporate. And that really shows that the product sticks. They're using it daily. To give you an idea, we are currently, I think, managing a small hundred thousand cloud resources through these customers. Yeah.

14:35Ben Murray: Okay. In the talk pricing, so now that you're GA, and we know in this new world and AI products, that pricing is all over the place. And is it usage? Is it subscription? Is it a mix? So how, and we know pricing changes, so I'm sure your pricing will iterate over time. But now that you're GA, how are you thinking about the pricing model? What kind of pricing model will you offer? So we have a three-tier pricing model, and that's really based on the core metric. There is the number of resources, number of cloud resources that you have. And of course, there is a limit on the number of tokens. the higher segment subscriptions have a higher limit on the number of tokens that can be used and the number of resources that can be managed etc and the number of architectures that can be created so that's in essence the base of the pricing but indeed prices change we have also noticed of course a spike in token costs but for now i think that we are testing the pricing model with many deals now in the pipeline and we are going to iterate on it and see whether it works but the pricing based on the number of resources, it looks like it's going to stick.

15:44Ben Murray: And Kim, it's great having you on as the CTO. So I don't know if Marty is coming up with these pricing models and then comes to you and, hey, here's how we're pricing. And so does this make your life hard, easy when, say, the commercial team comes up with, we're going to do, say, usage, we're going to do subscription, and we've got to track tokens? Like, how do you think about that as a CTO to be able to architect that so you know all the moving levers to determine, hey, are we actually going to make money with these pricing models? Yeah, I think this is a great question, Ryan. This is something that we thought of quite early.

16:21So we've been tracking metrics very quickly on and around, what is it costing us to run these environments? Especially because Dongart is processing some extremely sensitive information. we have a system that actually runs at higher cost, at higher base cost, to be able to safeguard the customer data to the best of our ability. So that means that there's a bit of a pricing cliff that you have, right? Maybe contrary to some of the multi-tenant SaaS applications that you see. So we have taken that into account from the start to make sure that we develop a pricing model that makes sense with the actual cost that we have so that we can always make money and make sure that it's still ritual for a customer, right?

17:02So we've built a system in a way that we comprise it in a way that we feel makes sense.

17:08Ben Murray: And we've seen a lot of these large tech companies shift from, say, they're offering an AI product line on a subscription basis. And then like we saw GitHub, and then it switched as a platform plus usage. And maybe all the heavy users were burning too much tokens on that subscription plan. So it sounds like you've got a three-tier structure. Sure. So are you, and there's a lot of theory going around, especially as like a CFO of what I want to track, but are you able to then track usage by customer and say tokens that are being consumed by your product line by customer? So you would know who are my power users and who are my light users?

17:45Ben Murray: And ultimately for me, how that impacts margins. Yeah, absolutely. So we can fully track per customer exactly what the costs are and what the token usage is. And the way it works is that we work with overach fees, right? So if you have say 20 million tokens or 200 million or 2 billion, whatever tokens you use more, you do pay extra for. So it is subscription plus usage. And that mainly comes down to the point that you made earlier, right? Tokens are expensive, especially for high reasoning models. So that's something that you need to take into account because if you don't price that in a smart way, you could really run yourself into issues.

18:20Ben Murray: It's great to hear that. Yeah, you can architect that way. You can track by customer, which seems like just a must have. And as far as the models that you're using, is this a type of product where you need those latest high reasoning models or can it be an older, cheaper model or where do you lean there? So a little bit of both, right? So there are many workloads that do not require frontier models. And the reason for that is, is that we chose to build what we call microagentic swarms, which are agents that have a very small area of responsibility. And you have to imagine the smaller the area of responsibility, the smaller the model can be.

18:56So we've done optimization there. But what we do see is that for some complex tasks, like orchestrating these swarms, you do want heavier models. So we have a mixture and we determine based on the workload what model we need in real time.

19:10Ben Murray: Okay, interesting. And you called that micro-agentic swarms? Yes. Okay. Okay, interesting. Yeah. Love that. Yeah. Okay. Really interesting because we hear a lot about model switching and frontier models versus the older models and, you know, picking the right model, especially for all the internal usage of AI. So really interesting. And so guys, at this stage of the business, really interesting, both co-founders, maybe different perspectives as a CEO, CTO, but I'm curious from both of you guys, what's a favorite number or metric, operational metric that you're focused on right now to guide the business?

19:42Ben Murray: Smada, do you want to kick it off? Because ultimately, those are the usual metrics you know, ARR, MRR, what have you, retention, the numbers. But at this stage, I'm really interested in the number of cloud resources that we manage, but also, for example, number of architecture that we create. Yeah. Because we are quite early in essence. We have some revenue, but we are in this big revenue. So those are the metrics that really shows how the platform is being used as we scale. Yeah. Interesting. And Kim, what are some of the numbers that you like to look at or maybe hand or keep you up at night.

20:18Yeah, so for me, it's, as you said, a double-edged sword. So I love to look at token spend. So how many tokens are we burning across these customers and how is that developing over time? So last month we saw north of 4 billion tokens being used by customers, which is amazing, right? We're seeing that the product is actively being used that it's delivering value and that customers are increasing their usage. But we also see, of course, cost rising so that is the double-edged sword that i mentioned there right because yeah that is not

20:50Ben Murray: cheap no not cheap but yeah four billion tokens a lot of tokens and i'm guessing that's only going to go up as as you guys scale so really interesting yeah and a couple stats of figures before we wrap up today and maddie i think you mentioned earlier about 30 staff and what's the composition is that heavy engineering right now or where does that staff fall yeah so we are heavily engineering i I think around 21 people engineering, different teams. Kim can expound on it, including a dedicated research team, machine learning scientists, developing a further technical mode. And then we have two people already in sales, two people in customer success, a person, a head of HR.

21:30We have since recently a product marketing manager. And then we have a CISO, of course, as part of the management team, very important. And also part of the management team is our CEO, Chief Operations Officer. Yeah, but mostly engineering.

21:46Ben Murray: Mostly. So Kim's controlling most of the budget right now as far as headcount goes. Hey Kim, with that budget, so mostly engineering, tell us a little bit about the Gentic side of coding. With each engineer, are they using, was there an approved Gentic coding tool? Tell us a little bit about your philosophy of AI coding. Yeah, first off, we try to give our people the least and greatest because they are asking for it. And what we're seeing is that the usage of AI has skyrocketed, right? So most of the code that we build is AI assisted or AI native, native regenerated. So what we do is we build extensive harnesses for our teams to make sure that we can govern how models are used and make sure that the outputs are good.

22:30So we use harnesses that include AI evals that include good guard railing observability, but we don't limit the budget of the engineers on token spend. So we see engineers spending significant amounts of their salary, if you will, on token spend. And funnily enough, we also see this from the operations side. So even on our sales team, we are spending a lot of money on tokens, but we equip them with the tools to do the job that they need to do. So right now we are looking heavily in how can we make that experience as seamless as possible? Now, can we equip people with a good way to work with AI without them all having to be experts and how to use AI?

23:12Okay.

23:13Ben Murray: And then there are there other tools or AI tools or processes where you can do quality checks on, say, that coding or the AI coding so you know it's legit? Yeah, absolutely. So we use a variety of tools as part of our engineering harness that does real-time checks on what we generate with the LLM. But we also use our own product. So we have a model context protocol server for Donguard that we inject in our development lifecycle that then also checks the output of the code that we generate. So we actually use our own product and our own development as well. Oh, that's great. Yeah, I appreciate that insight.

23:45Ben Murray: I was curious as CFO, what you're doing there. Guys, appreciate the time. Just one more question for you each. You're about a year in. We're halfway through 2026. So what, it'll be interesting, CEO versus CTO. So what's top of mind for you this year? What's really going through your thoughts? What keeps you up at night thinking about executing on the back half of the year? Matty, you want to go first? I have it, of course. Yeah. customers, but the pipeline looks very good. And we are signing, actually signing quite some partnerships with large IT companies that are going to adapt our platform to provide better services.

24:21That's something that has my focus, but I don't sleep less about it. Not so worried.

24:27Ben Murray: Yeah, still sleeping okay, but revenue here are always top of mind. Maybe some sleepless nights, who knows. But Kim, what's top of mind for you? Yeah, so the things that I'm getting really excited about and really focusing on is how can we incorporate these newer technologies like neurosymbolic AI world models as part of our product. And that's what Marios mentioned. We set up a new research team for to investigate these technologies and incorporate them in the product as soon as we can to really benefit from those new inventions as well. Okay. I'm going to have to Google that one. I have no clue what that is, but sounds interesting.

25:00Ben Murray: But that's why we have CTOs, right? Figure that stuff out. Guys, really appreciate your time and sharing your journey so far. if listeners would like to learn more about Dawnguard where should we send them online? Yeah, I'd say https colon slash slash Dawnguard dot AI Alright, perfect. Dawnguard dot AI so if you'd like to learn more about what Kim and Madi are doing, check out Dawnguard dot AI. And guys, really appreciate the time and sharing your journey. Thank you, Ben. Thanks for having us. Thanks.