Sign in Subscribe
Western Hemisphere

RWX Raises $12M Series A to Expand Its Dev Cloud for AI-Driven Engineering

The Editor

The Editor

— 4 min read
Share
RWX CEO Dan Manges
RWX CEO Dan Manges

RWX, a Columbus, Ohio company building cloud infrastructure for testing and validating software written by AI coding agents, has raised a US$12 million Series A led by Hyde Park Venture Partners. Existing investor Quiet Capital participated alongside The O.H.I.O. Fund, R1 Capital, and DV, with angel investment from Kohsuke Kawaguchi, creator of the Jenkins automation server, entrepreneur Bryan Johnson, and developer Theo Browne. Founded in 2022, RWX previously raised a US$7 million seed round led by Quiet Capital, bringing total funding to US$19 million.

The company was co-founded by CEO Dan Manges, who was founding chief technology officer at Braintree before its roughly US$800 million sale to PayPal and later co-founded Root Insurance, which went public at a valuation near US$6 billion, alongside CTO Tommy Graves. RWX's platform lets AI agents run tests without committing to Git, uses content-based caching to skip test runs that are not needed, sizes virtual machines per task, allows retrying individual steps rather than restarting entire pipelines, and provides sandboxes for agent environments, preview apps, and detection of unreliable tests. Customers include Honeycomb, Verkada, nCino, and Coalesce; Tropic reported cutting continuous integration time by 67% after migrating from GitHub Actions. "The limiting factor in software engineering is no longer writing code. It's validating it," Manges said. "Every major shift in software development has required new infrastructure, and AI-driven engineering is no different." The funding will go toward hiring and platform development.

Market Context

The thesis rests on a change in proportions rather than a new problem. When a developer spent two hours writing a feature, a ten-minute continuous integration run was a rounding error; when an agent produces the same code in five minutes, that same CI run can consume a majority of the cycle. Survey data supports the shift: GitLab's 2026 AI Accountability Report found 85% of respondents agree AI has moved the bottleneck from writing code to reviewing it, and Faros AI research covering 10,000 developers found that teams with high AI adoption merge 98% more pull requests while review time rises 91%. A Cursor study similarly found companies merging 39% more pull requests after adopting its coding agent.

The market RWX is entering is expanding accordingly, with AI coding tools projected to grow from roughly US$7.37 billion in 2025 to about US$30 billion by 2031, a compound annual growth rate above 26%, with DevOps and CI/CD among the faster-moving segments. RWX's positioning is deliberately adjacent to the crowded code-generation market: rather than competing with agents that write code, it sells the infrastructure those agents need to prove the code works. Hyde Park partner Greg Barnes said the founders "recognized early that AI-driven engineering would require a fundamentally different infrastructure stack."

The Signal

"The limiting factor in software engineering is no longer writing code. It's validating it." — Dan Manges, co-founder and CEO, RWX

Regional Relevance

For the United States: RWX is building in Columbus, Ohio, not Silicon Valley, and its investor list reflects that, with The O.H.I.O. Fund, co-founded by veteran investor Mark Kvamme, backing a developer infrastructure company competing directly against tools built on the coasts. Manges's track record, Braintree's sale to PayPal and Root's IPO, is unusual for the Midwest and helps explain how a Columbus company attracts angels of the caliber of the Jenkins creator. For US enterprises adopting AI coding agents, the practical question the company addresses is increasingly urgent: development velocity gains evaporate if verification infrastructure cannot keep pace.

For the global developer tools market: Continuous integration is a category long dominated by GitHub Actions, Jenkins, CircleCI, and cloud-provider offerings, used by engineering teams worldwide regardless of where the vendor sits. If AI agents genuinely require different CI architecture, as RWX argues, the shift would open one of the more entrenched segments of developer tooling to new entrants for the first time in years, with implications for teams from Bangalore to Berlin that currently run on incumbent pipelines.

The Other Side

Is validation really the bottleneck, or does speeding up CI just move the pressure downstream? Analysts have argued that the true constraint is deployment batching rather than review, noting that over 90% of teams ship changes in batches and that raising the rate of changes passing through review simply transfers pressure to manual approvals and release processes; on that reading, faster CI creates larger hidden queues rather than delivering value to users sooner.

Can a US$12 million Series A displace incumbents embedded in developers' daily workflow? GitHub Actions is free for many teams and lives inside the repository where code already sits, and Jenkins, CircleCI, and the major cloud providers all have entrenched positions; RWX's named customers are credible but few, and switching CI providers means rebuilding pipelines that engineering teams generally prefer not to touch.

What happens if the AI coding agents build validation themselves? The companies producing coding agents have both the incentive and the model access to fold testing and verification into their own products, and RWX's differentiation, agents testing without Git commits and caching that skips redundant runs, is architectural rather than proprietary in any obvious way, leaving open the question of how defensible the position is as agent platforms expand their scope.

Sources & Transparency

Read more