Agent Cloud Startup Daytona Secures $1B Valuation Backed by Industry Heavyweights

If you have spent any time deploying autonomous software loops or multi-step LLM workflows recently, you already know the dirty secret of modern cloud infrastructure: it was never built for autonomous machines. Traditional cloud services like AWS EC2 or Google Cloud Engine were architected decades ago for long-lived, predictable server applications managed by humans. When you unleash hundreds of autonomous AI agents that suddenly spin up thousands of short, stateful jobs completely unannounced, standard cloud environments choke, bleed resources, and rack up unpredictable bills.

Enter Daytona Platforms. Founded three years ago by CEO Ivan Burazin, alongside chief architect Goran Draganić and CTO Vedran Jukic, the New York-based startup announced on Friday that it has secured a $60 million Series B round at a $1 billion valuation. Led by FirstMark Growth—which also spearheaded the company's $24 million Series A in February—Daytona has now pulled in a cumulative $91 million in total funding. More importantly, the round has attracted a heavy-hitting roster of angel investors including Stripe’s Patrick Collison, Shopify’s Tobi Lütke, Perplexity’s Aravind Srinivas, Zoom’s Eric Yuan, Datadog’s Olivier Pomel, Quora and OpenAI board member Adam D’Angelo, and Mercor’s Brendan Foody, alongside institutional participation from Qualcomm Ventures, Samsung Next, and Okta Ventures.

src="https://images.unsplash.com/photo-1618005182384-a83a8bd57fbe?auto=format&fit=crop&w=1200&q=80" alt="Abstract representation of cloud infrastructure and computing networks"

Daytona aims to solve the infrastructure bottleneck for autonomous software agents by creating dedicated execution environments.

Giving Every Autonomous Agent Its Own Computer

The core thesis behind Daytona is deceptively simple: treat every single agent as an independent user requiring its own dedicated machine. In a blog post detailing the new financing, founder Ivan Burazin explained that the startup’s infrastructure functions similarly to providing every software agent with a personal computer loaded with sandboxed versions of Linux, Windows, macOS, and dedicated GPU access.

Unlike traditional virtual machines that require manual configuration and continuous monitoring, Daytona's cloud layer is designed specifically to handle the erratic, bursty nature of AI agents. When an agent needs to compile code, scrape live web data, run tests, or execute terminal commands, it spawns an isolated workspace instantly and destroys it just as fast when the task concludes. This eliminates the security vulnerabilities of running untrusted agent-generated code on shared internal networks while matching the exact performance profile required for long-running autonomous operations.

Backing from the Architectures of the Modern Internet

The speed at which Daytona has scaled to unicorn status reflects a broader industry panic regarding agent orchestration. While foundational model providers focus heavily on reasoning capabilities and token windows, enterprise developers consistently hit a brick wall when trying to deploy these models into production environments that require persistent file systems, secure network sandboxing, and scalable compute.

By bringing together executives and founders from Stripe, Shopify, Perplexity, and Datadog, Daytona has positioned itself not just as a piece of developer tooling, but as a foundational infrastructure layer for the upcoming wave of autonomous enterprise software. The participation of corporate venture arms like Qualcomm Ventures and Samsung Next further signals that hardware manufacturers are keenly interested in how edge devices and silicon will interface with agent-heavy cloud workloads.

Next Steps and Production Implementation

With $60 million in fresh capital now in the bank, Daytona plans to accelerate the rollout of its multi-OS sandboxing architecture and expand enterprise integrations. Developers and engineering teams looking to test long-running agents without managing raw Kubernetes clusters or custom Docker containers can access Daytona's platform directly via their official developer portal. What remains to be seen is how traditional hyperscalers like Amazon and Microsoft will respond to specialized startups carving out dedicated infrastructure niches specifically engineered for non-human workloads.