Croatian infrastructure startup, founded by Ivan Burazin, Vedran Jukić, and Goran Draganić, closes a round led by FirstMark Capital. Daytona provides sandboxes — programmable environments where AI agents execute code, fork workflows, and scale to millions of concurrent instances.
Daytona announced today it has raised €20.4 ($24) million in Series A funding. The round was led by FirstMark Capital, the New York venture firm that backed Airbnb, Pinterest, Shopify, Discord, and Riot Games in their early stages. Previous investors Pace Capital, Upfront Ventures, Darkmode, and E2VC participated, alongside strategic investments from Datadog and Figma Ventures, while angels include founders from Fal, T3, Factory.ai, and Neon.
The round brings Daytona’s total raised to $31 million since its 2023 founding and follows a $5 million seed led by Upfront Ventures in June 2024.
The funding addresses a structural shift in cloud computing
AI agents require fundamentally different computing environments than traditional applications. Current cloud systems were designed for stateless, immutable production workloads that run the same code the same way every time. Agents need real computers where they can install packages, run experiments, fork execution paths, and recover from failures.
Croatian Daytona solves this through what it calls sandboxes: programmatic, composable computers that launch in milliseconds. These are full computing environments where CPU, memory, storage, GPU, networking, and the operating system can be composed on demand, then started, paused, forked, snapshotted, or destroyed at any point.
“We believe the next infrastructure shift is from human-centric cloud primitives to agent-native ones,” said Matt Turck, partner at FirstMark and new member of Daytona’s board.
“Daytona’s breakthrough is making ‘a computer for every agent’ practical: instant startup, persistent state, and the tooling agents need to write code, use Git, and execute safely at scale. That’s a foundational building block for the agentic economy, and we are thrilled to partner with Ivan, Vedran and the Daytona team.”
Their recent tracktion is a clear proof of this shift. Daytona reports it reached $1 million in forward annual recurring revenue in under three months. Six weeks later, it doubled. Customers now span early-stage YC companies and Fortune 100 enterprises, including LangChain, Turing, Writer, and SambaNova. Clients like them use the platform for three core cases: code execution, computer use, and reinforcement learning.
How it all started?
Daytona started as an open-source development environment manager in 2023. Founder and CEO Ivan Burazin, previously Chief Developer at Croatian unicorn Infobip, built the company alongside Goran Draganić and Vedran Jukić to solve a persistent problem: developers lose 56 percent of productive time to inefficient environments, according to IEEE Journal research.
The team’s first product automated environment setups, enabling developers to spin up working environments with a single command. According to the company’s data, by late 2023, Daytona had nearly doubled its ARR.
The shift to agent infrastructure came as generative AI evolved from chatbots to autonomous systems attempting real tasks. “The biggest change is the shift from Chatbot AI to Agentic AI,” Burazin explained for The Recursive.
“While earlier models just generated text, modern AI agents now attempt to perform tasks. However, this is where we encounter a key obstacle: for an agent to truly work as a digital expert, they need a ‘workspace’: a computer, an operating system, and a secure environment.”
Daytona pivoted to provide exactly that — standardized, isolated sandboxes where agents write code, test applications, and execute operations without risking customer infrastructure.
Croatian technical center
Daytona today operates from New York headquarters, serves key customers in San Francisco, yet runs its 20-person technical team primarily from offices in Split and Zagreb, Croatia.
Given that they are reaching the scaleup level, we were interested to find out what structural and strategic changes await their business and team. First thing, the new funding will expand capacity and accelerate international growth, with particular focus on the US and Western European markets. In other instances:
“We are introducing security and operational certificates (such as SOC2), which is necessary for working with clients from the banking and healthcare sectors. Also, we will focus on making Daytona so easy to set up that teams within large organizations can start using the tool without months of implementations.”
Kevin Zhang, General Partner at Upfront Ventures and Daytona board member, emphasized the team’s execution speed: “The team’s relentless pace and obsession over developer experience have been truly inspiring to witness, best reflected by the constant activity and customer love in the Daytona Slack and X. It is just the beginning of this new agent era infrastructure opportunity, and we cannot think of a better and more focused team than Daytona to take it on.”
Upfront led Daytona’s $5 million seed eight months ago, joined by existing investor 500 Global and angels including the founders of Postman, Netlify, and Supabase.
Investors able to support Daytona’s scaling
The investor roster reflects both enterprise credibility and developer-tool expertise. FirstMark Capital’s track record includes early bets on companies that became standards in their categories. On the other hand, strategic investors like Datadog and Figma Ventures bring operational insight into scaling developer infrastructure.
Beyond the lead investors, the round includes angel investors with operational backgrounds: Gorkem Yurtseven, co-founder of Fal; Theo Browne, founder of T3 Chat; Eno Reyes, co-founder of Factory.ai; and Nikita Shamgunov, founder of Neon. “We’re grateful for their early belief and for the founder perspective they bring as we build agent-native infrastructure,” the company said.
Daytona positions sandboxes as a starting point. As agents take on more work, the company argues the entire infrastructure stack will need rebuilding from first principles with the agent as the primary consumer.
