Founded by veterans of Scale AI, Google and Stripe, Runloop is helping companies automate evaluation and get their AI coding agents deployed up to six months faster
Runloop, the only enterprise-grade infrastructure platform that enables the development, evaluation and scalable deployment of AI coding agents, announced today that it has raised a $7M seed round led by The General Partnership with participation from Blank Ventures. Runloop will use the funds to accelerate hiring and delivery on its product roadmap to leverage strong demand for its AI coding agent deployment and evaluation platform.
“AI coding agents are already widely used, but there’s a critical gap between prototypes and production,” said Dan Portillo, co-founder at The General Partnership. “Any company looking to deploy an autonomous AI coding agent needs a solution like Runloop. We think this approach will be ubiquitous among dev teams by the end of 2025.” This insight has already been proven out by the recent announcements of OpenAI Codex, Cursor background agents and Google Jules.
“AI coding agents are the future but they need developer tools that are distinct from those of human developers. Providing that richly tooled environment along with the evaluation mechanisms required for effective deployment is Runloop’s mission,” said Jonathan Wall, co-founder and CEO of Runloop. “We help AI coding agents get into production in a fraction of the time.”
Deploying AI coding agents in production is incredibly challenging. Runloop provides secure and isolated sandboxes (called Runloop devboxes) for developers to create, run and evaluate their models in. Runloop offers comprehensive tooling to support the overall developer experience with features like direct GitHub repository integration, snapshots and blueprints to ease every step when deploying agents.
Evaluating these AI coding agents has typically been a fragmented process that requires multiple tools. Many companies still do it manually. Runloop’s Public Benchmarks, provides organizations with on-demand access to industry-standard performance testing for AI coding agents. Benchmark results can be used internally for model improvement or shared to demonstrate model quality externally.
Runloop was founded by a group of developers from Stripe led by Wall, who recognized that the impending wave of AI coding agents would require scalable infrastructure and evaluation frameworks to ensure global use of coding agents are possible. Wall was previously co-founder of Google Wallet and brought tap-to-pay technology to daily use in the US. After leaving Google, he co-founded fintech startup Index which was then acquired by Stripe.
Runloop customer Dan Robinson, CEO of Detail.dev, said, “Runloop has been killer for our business. We couldn’t have gotten to market so quickly without it. Instead of burning months building infrastructure, we’ve been able to focus on what we’re passionate about: creating agents that crush tech debt. Obvious choice to bridge the infra gap between ‘cool demo that runs locally’ and an AI devtool that can scale. Runloop basically compressed our go-to-market timeline by six months.”
Headquartered in San Francisco, Runloop has a team of 12 and is growing quickly. Team members are from well-known companies such as Vercel, ScaleAI, Google and Stripe.