Technology funding: infrastructure for autonomous AI agents
Developers and AI product teams are starting to pay for the same thing they already buy in cloud and DevOps: a reliable way to run code safely at scale. Daytona is positioning itself to sell that workflow for autonomous AI agents, removing the pain of running untrusted or variable agent code in controlled, reproducible environments.
Croatian-founded, Croatian-American startup Daytona has raised $60 million in a Series B funding round, equal to ~EUR 56m. The round was led by FirstMark Growth and included Qualcomm Ventures, Samsung Next, and Okta Ventures, according to multiple reports. Daytona said the financing valued the company at $1 billion, making it a unicorn.
The company’s total funding has reached approximately $91 million. Multiple reports also describe Daytona as Croatia’s third unicorn, a milestone that underlines how much of today’s AI infrastructure capital is being deployed from outside smaller European markets.
From dev tooling to agent execution infrastructure
Daytona has shifted from development-environment tooling toward sandboxed execution environments for AI agents. That repositioning matters because it moves Daytona from a developer convenience category into a higher-stakes infrastructure layer, where buyers care about isolation, compliance, reliability, and cost control.
The company markets its Agent Cloud as infrastructure for autonomous AI agents. The product supports isolated environments across Linux, Windows, macOS, and GPU workloads, suggesting Daytona is aiming at a broad “agent-execution infrastructure” category rather than a narrow IDE or dev environment niche.
In practical terms, this kind of platform is designed to help teams run agent workloads that:
- Need clean, reproducible environments per task or per customer
- Must execute safely with strong isolation boundaries
- Span heterogeneous stacks (different OS requirements, GPU-enabled workloads)
If Daytona can become the default runtime layer for agents, it can embed deeply into production workflows. That implementation depth is a key retention lever in infrastructure markets.
Why this round fits the current AI infrastructure playbook
This financing is a with-trend signal: investors are backing picks-and-shovels infrastructure that makes agentic AI deployable in real-world environments, not just demos.
The syndicate also stands out. Alongside institutional and strategic investors, the round included senior founders and operators such as Patrick Collison, Tobi Lütke, Aravind Srinivas, Eric Yuan, Olivier Pomel, Adam D’Angelo, and Brendan Foody. The presence of these technology leaders, together with Qualcomm Ventures, Samsung Next, and Okta Ventures, signals confidence that the “agent runtime” layer will become a meaningful budget line over time.
Strategic participation can matter commercially if it translates into:
- Distribution channels and ecosystem integrations
- Design partnerships that harden the product for enterprise requirements
- Credibility in long enterprise sales cycles where runtime and security are scrutinised
Geographic signal: offshore success, global capital
Daytona’s Series B was led by New York-based FirstMark, with a heavy U.S. and global strategic investor mix. That reinforces a familiar pattern for European and CEE-origin companies in frontier infrastructure categories: R&D and founding talent can be local, but late-stage capital and go-to-market networks often concentrate in the U.S.
For Croatia specifically, being labelled a third unicorn is symbolic, but the more operational takeaway is that globally competitive AI infrastructure companies can emerge from smaller ecosystems when they attach early to international capital and customer bases.
What this enables
- Faster scaling of Agent Cloud as a cross-OS, GPU-capable execution layer for AI agents
- Deeper enterprise readiness through hardening around isolation, governance, and reliability requirements
- More integrations and partnerships driven by the presence of strategic investors (inference)
What to watch
- Category definition: whether “agent execution infrastructure” becomes a distinct budget and buyer persona
- Go-to-market motion: self-serve developer adoption versus enterprise platform sales
- Moat and switching costs: how tightly Daytona can embed into production agent workflows
- Strategic pull-through: whether investor relationships convert into integrations, customers, or distribution