A small Prague startup just landed a fresh round of cash to fix one of the most painful problems in software today. Zerops, a Platform-as-a-Service startup redesigning cloud architecture, has raised a $2 million seed round led by Gi21 Capital to remove the separation between development and production environments. The deal arrives at a moment when AI coding agents are flooding into developer workflows and breaking old infrastructure assumptions wide open.

Inside the $2M Seed Round Backing Zerops

The funding pushes Zerops into a new phase. The Prague-based platform-as-a-service startup raised a €1.7 million ($2 million) seed round led by Gi21 Capital to expand its global infrastructure in the US and Asia, accelerate product development, and grow its team.

It is the company’s second outside check. In 2024, Zerops secured a $500,000 pre-seed round from Presto Ventures and Gi21 Capital to build deeper integrations with popular frameworks and open-source software.

The new round will not just sit in a bank account. The investment will primarily support distribution efforts, including developer marketing, open-source sponsorships, hackathons, conference participation, and creator partnerships, while the team currently has 12 members, with two more hires in progress.

Zerops AI cloud platform seed funding announcement

“We’re moving from millions of developers to millions of developers working alongside AI agents. Most platforms weren’t designed for either of these changes. Zerops was.”
Damir Špoljarič, founder of Gi21 Capital

Closing the Gap Between Dev and Production

The pitch behind Zerops is simple but bold. It removes the separation between development and production environments, a long-standing flaw in cloud architecture that causes deployment failures for developers and AI coding agents alike, creating a unified environment where applications behave identically from development through to production.

That gap has cost teams countless late nights. On Zerops, there are no environment tiers; applications run within a single project where code behaves the same way, regardless of scale. Because the infrastructure is consistent from the start, deploying a production-ready system requires a single click, not weeks of configuration.

CEO Aleš Rechtorík frames the change in plain language. “Most platforms ask you to trust that development and production are close enough. We removed the gap entirely by rethinking how cloud architecture should work from the ground up,” says Rechtorík.

The cost of broken environments is no small line item. Gartner forecasts that global cloud expenditure will hit $723 billion by 2025, yet developers devote just 16% of their time to coding, while the remaining effort is spent on infrastructure and tools.

New Control Panel Built for AI Coding Agents

The biggest news inside the news is a product, not a number. The company is also introducing Zerops Control Panel (ZCP), a feature designed for AI-driven development that connects AI coding agents, like Claude, Codex, or Gemini, directly to real cloud infrastructure inside a Zerops project, allowing them to build, deploy, and debug applications in real conditions rather than isolated environments.

This is where the funding round meets a very current pain point. As AI coding tools become mainstream, the challenge grows: 45% of developers report that debugging AI-generated code takes longer than writing it, mainly because AI operates in sandboxed environments that do not reflect real systems.

Rechtorík calls the shift a reset for cloud thinking. “AWS solved one problem but created another, you just swapped DevOps engineers for AWS experts. PaaS platforms then overcorrected, becoming black boxes. We’re going back to the middle,” he says.

Key features of the Zerops Control Panel:

• Direct integration with Claude, Codex and Gemini coding agents
• Live cloud infrastructure access instead of sandboxed dev boxes
• Same workspace for AI output and human review
• Production-ready deployment from the first run

Bare-Metal Edge and Global Expansion Plans

Where Zerops really tries to break from the pack is the hardware layer. Unlike most PaaS competitors like Render, Fly.io, and Railway, Zerops runs on its own bare-metal infrastructure rather than renting capacity from AWS, and this cost structure lets it price up to four times lower than legacy cloud platforms.

Developers get more than savings. The platform runs applications in full Linux containers, not restricted app containers, giving developers the same level of access as on their own machines, including real-time visibility and control over running processes, and includes more than 15 built-in services such as databases, search engines, and messaging systems, reducing the need for external integrations compared to the typical two or three offered by most platforms.

The expansion map is also taking shape fast. Zerops operates a European data centre, recently launched a US East region, and plans to open US West and Singapore data centres in the coming months.

What This Round Means for the Future of Cloud Building

The deal lands in a market that is shifting under everyone’s feet. Rising bills and rising AI agent usage are forcing platforms to rethink the basics, and the founders see this as their window.

The roots of Zerops also help explain its confidence. Rechtorík and Jan Saidl developed Zerops as the lead team at Via Hosting, a Czech provider supporting infrastructure for about half of the Czech and Slovak e-commerce. After the project was sold to German cloud company Contabo, it was discontinued due to corporate restructuring, and Rechtorík later reacquired the project and spent over a year fundraising during the post-2022 VC downturn before relaunching it independently at the end of 2024.

For developers tired of the same broken loop of staging, surprise crashes and rewrites, this story carries real hope. The promise is not flashy. It is simply that what they build today will keep working tomorrow, even as AI agents start writing more of the code beside them. Share your thoughts in the comments. Do you trust AI agents to deploy straight to production, or does the idea still make you nervous.