Dash0 raises $110M Series B at $1B valuation: Mirko Novakovic, CEO at Dash0

0:00Mirko Novakovic: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 Cibesia 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.

0:32Mirko Novakovic:Hello and welcome back to the Scaling Europe show presented by Deal. I'm Seb Johnson. Today I'm joined by an amazing guest, Merkur, CEO of DashLever, who's got some amazing news. Take it away. Yeah, hi. Good to be here. Yeah, and we are announcing our Series B. It's a$100 million round on a billion valuation led by Balderton and also from our existing investors, Accel, Cherry and Dig. Amazing. That's absolutely amazing news. And you've become a unicorn. That must be like an amazing milestone to hit. And I'm sure there's a lot of work to do. But talk to me a bit about the rounds. You said Bolton have led.

1:07Mirko Novakovic:What gave them the conviction to invest, you know, a reasonably large ticket size at a great valuation? Yeah, I think we raised our Series A only five months ago. And then we essentially crushed last year from our sales numbers. So we came out 5x higher than we had planned. and then we also had a really good start into Q1. We crushed through 10 million in ARR, growing pretty fast. We want to grow by 10x this year. And so I took the opportunity to go out and fundraise to fuel that growth, right? Because we are expanding our sales team in the US heavily, in Europe, but also invest in our product, which we can talk about the full shift to agentic and AI is amazing what's happening right now, also in our space.

2:02And so that's how we went out. We got a lot of traction on this fundraise and then ended up with Rana from Balderton leading the round. And yeah, so we are super excited about it.

2:13Mirko Novakovic:Amazing. Where's that growth come from? You know, what enabled you to smash into the last year so aggressively? It is also, so look, I'm 26 years in the observability space. And when I started DevZero, I was convinced that there is a market. So it's a very competitive market, kind of a red ocean market with some large players. So I was actually not expecting us to be able to grow that fast. And the reason why it happened is essentially that it looks like the current market is broken. A lot of the customers we are talking to, they generate a lot of data. So in our space, you basically pay for the amount of data you send to the vendor.

2:54number of logs or metrics or gigabytes and that number is exploding and so is the cost for the tools and and by having more data that doesn't mean that it's better for you I always say our job is to find the needle in the haystack and if the haystack is bigger it doesn't make it easier to find the needle right and if you then also pay for the hay it's it's not not the right way of doing it, right? So that's essentially where we are in. And so what we did is we helped customers reduce the amount of data and we helped them finding needle in Haystack by applying AI on it. And we're using an open standard called open telemetry, which makes lock-in way less than with proprietary technology.

3:44Mirko Novakovic:Interesting. Okay. And so let's touch about this move to agentic technology then like how how has that changed and how has that enabled the business to kind of accelerate even faster i mean first of all i have to be very honest the way ai is changing at the moment and the rate of change is is is happening so fast that it's really hard to have a clear strategy road level in it right so what we are trying at the moment is really keeping up with the pace. If you look at, if you would have asked me, I don't know, six months ago about coding agents, right? I would have not expected that 90 % or more of our code is generated by an agent today, but it is, right?

4:27Today, coding agents are generating the majority of our code. And that also changes the way how you use observability. So observability means we are monitoring production code, right? We are monitoring applications in production and we give developers and SREs the insights why something is slow or why you have errors, why something is broken or doesn't scale. And now think about this whole new way of developing code. An agent creates code. I talked to my CTO last week. He created a PR that was like 30 ,000 lines of code. it's almost impossible to review it so you push it into production and now we come into play the zero comes into play our platform because now we can our agents pick that new change up and we are looking at that change and we can investigate pretty quickly if that's working or not and then we can automatically roll back or we can put more users on it right so gradually rolling out the feature so it's kind of your insurance policy in production now that that you have with observability to to figure out if all the code that you are deploying now on a

5:40Mirko Novakovic:much higher frequency is actually up and running wow that's insane it's uh wow i mean yeah it must be um it must be great for the business as well like you've got a lot of the tech the way the market is moving and changing must be great for the business and so i guess like looking back to when you first started, Mirko. What did you get right that enabled you to take a big bet on Dash Zero that now seems to be being proven right by the way the market is moving? Again, I have to be honest, right? We were a little bit lucky that our architecture was built for this AI world. And when we started Dash Zero back in May 23, I haven't seen what I was just talking about, right?

6:25I haven't seen 90 % of the code being generated, higher frequency. I haven't seen that. So when we started, the basic idea was that we wanted to build an observability platform on top of that new standard open telemetry, which is an open source, open standard for telemetry data. And we did that. As the first platform, we were built from the ground up. All the data in our backend, the data you send us is open telemetry. and it turns out that all the LLMs out there are trained on it already because you train those LLMs on publicly available information and so all the LLMs, we use Cloud at the moment, so Cloud is fully trained on open telemetry.

7:09If you take an open telemetry trace and you paste it into Cloud or ChatGPT, it will understand what it is because it's a publicly available standard. It's fully documented. There is a semantic convention. All the agents who do the code instrumentation are open source. So essentially, it was super easy for us to apply LLMs on top of our architecture because all the data was directly understood by the LLM, right? Including the context. And the other thing we did from day one was we wanted to make it easy for the user to create context. Context means in a lot of situations, you have different types of data.

7:50You have a log, you have a trace, which is like how the code flows through your application, and you have a metric, could be something like the CPU usage of your server, and you want to have that in context to understand the problem. So think about something is slow, and then you want to understand, oh, could that be because I didn't have CPU? You need the CPU metric of the server that this code was running on. That's context, right? You don't need the CPU of all the thousand servers of your system. You only need that one where that code was executed. And to do that, you need context. And so we have built a system that has all the data always in context.

8:29And it turns out also that is very relevant for LLMs because their context window is limited. You can't give all the data to an LLM. So you have to give the right data with the right context to the LLM so that it can make sense out of it. And so all these features that we have integrated into our platform finally made total sense for agents. And again, not really built from day one with that thing in mind, but like a year ago, we really understood how this is changing and transforming our market. And since then, we are working full steam on that platform we call Agent Zero, which is our agentic platform on top of our observability platform.

9:13and it's amazing, right? The adoption rate is amazing. Customers love it. We really help solving real production problems. And now we are taking it a step further and we want to become the agentic platform for production data. So also customers can build their own agents on top of our platform. So very specialized agents with specialized context of your architecture, of your business, so that we can better understand what's working, what's not working.

9:42Mirko Novakovic:and you said customers love it i mean i think i bet you've got over 600 now yeah how how does usage how does retention how does that look like for your current customers yeah we have 150 net revenue retention so customers who start with us they really grow and they add more data they add more functionality and they consolidate their platforms on top of dash zero on top of open telemetry and our agentic platform. And yeah, and so basically almost every user today that logs into Dash Zero uses agent zero, uses AI to help making sense out of the data, right? Because it has also become so complex and the rate of change is so fast that it's almost impossible for a normal user to figure out the context and what's happening, right?

10:34So you basically need agents in production to help you figuring out where the problems are interesting and how has

10:41Mirko Novakovic:your customer profile changed you know you've got 600 accounts now have you seen a change and has was it very small clients and you're now you're now moving towards enterprise you see enterprise being a huge driver of growth in 2026 can you talk a bit about the customer profile and how that's changed yeah yeah yeah you are spot on so when we started we were mostly looking at smaller startups. We were fully product-like growths. So people logged into our platform, sent us data, put in a credit card and buy the product, right? And that's still the case. So we are generating 15 to 20 customers a week at the moment only via PLG.

11:17And that could be small customers or teams inside of larger customers just using the product and testing it and putting in a credit card. But then we also figured out that this problem is even more dominant in larger accounts. You can think of an SMB, maybe it's also painful to pay for an observability tool, have a lot of data, but these larger customers, so one of our largest customers is Alando, e-commerce player here in Germany. I mean, they have trillions of data points they are creating, trillions, right? It's amazing. And to help them, they really need a consolidated view of all the data, all their services in one platform and AI on top, helping them to make sense.

12:05They have hundreds or thousands of developers building things and needing the feedback out of production, how things are up and running and if there are potential problems.

12:18Mirko Novakovic:And let's talk about the raise itself. So you've raised$100 million. dollars you're gonna you're gonna go big on expansion where do you see that expansion going to because i know you've got a lot of customers in america are you doubling down do you think you're gonna see most of your growth then and you're gonna be expanding the team there what's that money going to be useful yeah so today we are around 130 people 100 of them in europe 30 in the us so we are having in europe at the moment and the market for observability the by far biggest market is the u.s market so we will expand so we have offices in boston and new york go to market office we have around 30 people um on the ground at the moment and and and we will expand that team heavily with that round and um yeah we we want to uh grow into into the u.s market pretty pretty quickly amazing and it seems like you're you're doing well already um how are you thinking about go to market you know is it classic b2b and you know are you thinking about any sort of tactics to get yourself in with the best accounts yeah i mean it's a combination right it's a combination of plg i love plg and i i i love plg because it's it keeps you honest right yeah um because there's basically no human interaction somebody comes in needs to figure out the product send you data two-week trial put in the credit card or not right this is kind kind of what what i really enjoy and that also drives pipeline right because sometimes you get a team in a larger account and now you can engage with that team and figure out if you can expand into those accounts so so yes for larger deals you always need enterprise sales right you can't make um i mean we we are working on seven figure, eight figure deals right now.

14:09So you can't do seven, eight figure deals yearly, right? Seven, eight figure deals without a sales team and without real engagement with the customer and also without things like forward deployed engineers, right? Because you need to help customers adopt your tool and get all the value out of it. And by the way, our pricing model is constructed in a way that if you don't have adoption, you don't pay, right? So it's also in our interest making sure that the customer adopts this product because otherwise we don't get paid, right? Because you only pay for consumption. You don't pay a license fee if you don't send us data.

14:48Mirko Novakovic:Yeah, it's a really interesting evolution of the classic SaaS model, right? Where it used to just be like paying for like a boner seat or an access. Now you really are incredibly motivated to get your customers extracting the most value and using the products as much as possible. Have you got lots of Ford to Ford engineers? Is that something you're investing in a lot? Yes, we are. We are. We call them solution architects. We haven't renamed it. So we have solution architects who are essentially there for the customer to help them adopt our tool. And very technical people, mostly having developer background so that they can really help the customers implement open telemetry, making sure that the collectors are configured the right way and reducing the amount of data so that's what they do amazing and looking forward you know you've got a big kind of like war chest now of capital to deploy to grow by the end of the year how what will you need to do to look back and say okay this year has been a massive success you know we've deployed the capital efficiently we're at where we want it to be i mean there are multiple things right i i think on the product side we want to heavily invest in our AI platform, Agent Zero, and we are building a lot of agentic functionality capabilities on top, including by the way things like security, because I think that those markets will finally converge, right, observability, security, we are operating on the same data.

16:16You can see things like product analytics also working very well with and user monitoring data and then having an agent figuring out if your usability is good or not. So I think we have a lot of ideas what we want to build with Agentic on top of the production data. That's one thing. So we will invest in R &D and product. And on the other hand, go-to-market is our priority, right? So hiring more salespeople, hiring more forward deployed engineers but also investing in marketing, right? I mean, marketing is changing also heavily with AI, right? So how you create content, how fast you have to be these days.

17:02You are a good example, right, of how the whole kind of content and marketing creation is changing in this world, right? So I think we have to adapt to that trend.

17:15Mirko Novakovic:Yeah, you've got to be very present on almost every channel. Exactly. And I want to ask one final question because we're running out of time. But when you look at the market, you look at the rise of AI, what companies do you think are going to be the AI companies that survive in five or 10 years? Which of the companies that are building really robust products that are very sticky and very durable? If I would know, right? I think I have to be really honest, but at the moment, I find it very hard to predict the next six months. Yeah. predicting five to ten years for me at the moment is almost impossible right at the moment I would say that the companies like entropic who really build foundational models that everyone is using in that area of code production data I I don't know but I think they have a very robust business model at the moment if you see how they scale right from 1 billion to 13 14 billion dollar in revenue billion, right?

18:17In a year. It's crazy. Crazy, right? Never seen before. I think very robust model, but also there, will there be an end, right? Will at some point all the models be equally good at everything? And can you then use a smaller open source model for a fraction of a price? I don't know, right? I think it's very hard to predict. At the moment, I would say companies like Entropic or OpenAI, They look very promising to me, right? Because they have an absolute incredible capability and it costs billions to train those models. So you can't just copy it at the moment. But five years from now, who knows, right?

18:58I don't know. Very hard to say.

19:00Mirko Novakovic:The rate of change is increasing. Yeah, I mean, yeah. Well, look, thank you so much for joining me. It's absolutely amazing news for you and the team. So yeah, congratulations and best of luck with it. Thank you, Seb. It was nice being here.

19:16Thank you.