Summary

• Namespace has raised a $42 million Series B led by Scale Venture Partners, taking total funding to $65 million.
• The company provides computing environments for coding, continuous integration, testing and AI development agents across Mac, Windows and Linux.
• Namespace says revenue has grown eightfold over the past year as AI coding increases demand for build and test infrastructure.

AI can now produce code faster than many engineering organisations can build, test and validate it, pushing a different part of software development towards the centre of the infrastructure market.

Zurich founded Namespace has raised $42 million in a Series B led by Scale Venture Partners, only seven months after completing its previous financing. NEA, 20VC, Essence, Burst Capital and other investors joined the round, taking total funding to $65 million.

The company provides computing infrastructure for coding, continuous integration, software builds and testing, including environments that AI coding agents can operate directly. Namespace supports Mac, Windows and Linux workloads and runs its own compute, storage, networking and scheduling stack rather than functioning only as a software layer over a general cloud platform.

Its timing reflects an increasingly obvious side effect of coding agents. Generating another software change may take seconds, but that code still has to be compiled, tested against existing systems, checked for regressions and prepared for deployment. Faster creation consequently increases the amount of downstream computing work rather than removing it.

Software generation changes where engineering waits

Traditional development workflows contain numerous pauses that have little to do with writing code because developers wait for test environments to start, dependencies to download, large repositories to check out and continuous integration jobs to reach available machines.

Those delays could already be expensive when engineers produced changes manually, but autonomous coding tools alter the volume involved. An agent can prepare another change while a previous one is still waiting for tests, allowing the rate of software generation to exceed the throughput of infrastructure designed around human coding speed.

Build and test capacity can therefore become a constraint on AI assisted development. Organisations investing heavily in coding agents can lose much of the productivity gain if each agent still has to queue for a slow shared runner or spend several minutes rebuilding the same development environment.

Namespace is trying to remove those delays by controlling more of the underlying infrastructure. Its platform provides temporary compute for development agents and continuous integration workloads, while storage and caching systems are intended to avoid repeatedly downloading the same repositories, dependencies and container images.

The company recently introduced Git Snapshots, which prepare a repository once and reuse that state across later Linux and macOS runners. Namespace says early customers have seen repository checkouts run up to 6.5 times faster than conventional approaches, although that figure comes from the company’s own measurements and will vary considerably between projects.

AI agents need machines, not only models

A chatbot suggesting code inside an editor can rely largely on the developer’s existing computer, whereas an autonomous agent expected to change a repository, install dependencies, run tests and inspect failures needs somewhere to execute those actions safely.

Namespace calls those environments Devboxes. Each gives an agent access to a complete virtual machine containing the required codebase, tools and test infrastructure rather than a narrowly constrained execution sandbox.

Greater capability brings operational requirements around isolation, credentials and cost because an agent able to build an application or interact with development services may need network access and secrets, making its compute environment part of the organisation’s security boundary.

Running many agents concurrently can also increase infrastructure consumption quickly. A developer might previously have triggered a handful of substantial builds during a working session, while several autonomous agents can generate and test alternatives continuously.

Developer infrastructure suppliers therefore have an incentive to sell efficiency alongside raw capacity. Faster startup, cache reuse and more granular compute allocation can reduce idle or duplicated work even as the number of automated development tasks rises.

Specialist clouds are moving into narrower workloads

Namespace’s strategy also reflects increasing specialisation inside cloud computing. The major hyperscalers provide broad infrastructure platforms, while smaller providers are optimising hardware and software around particular workloads where latency, performance or developer experience can justify a dedicated service.

Build and test jobs are unusually sensitive to single core performance, storage speed and startup latency because they run frequently and often for short periods. A modest improvement repeated across thousands of engineering jobs can therefore produce a meaningful reduction in aggregate waiting time.

Namespace says more than 1,000 companies use its technology and lists customers including Ramp, Framer and Vanta. Its current platform supports GitHub Actions as well as development environments for AI agents and provides Apple hardware for workflows requiring macOS.

The latest funding will support further product development and expansion of the company’s own compute footprint. That makes the business more capital intensive than many software companies because adding capacity requires physical infrastructure as well as engineers.

Revenue has reportedly grown eightfold over the past 12 months, although that remains a company supplied figure from an early growth stage where percentage increases can be large from a comparatively small base.

The more important test is whether AI coding continues to increase the total volume of software engineering work enough to sustain a distinct infrastructure category beneath it. Coding models are improving quickly and competition between them remains intense, but every generated change still needs somewhere to run.

If software creation becomes cheaper and faster, the systems responsible for proving that code actually works may become more valuable rather than less. Namespace is raising money on the assumption that the next bottleneck in AI development sits below the agent.