Open-source development platform builder Coder Technologies Inc. today is unveiling a suite of governance and execution capabilities designed to help enterprises integrate artificial intelligence coding agents into their software development lifecycles without compromising security, compliance or platform control.

The launch extends the company’s self-hosted developer-environment platform to a “full-stack foundation for governed AI development,” aimed at organizations struggling to adopt AI tools safely at scale.

Three new features — AI Bridge, Agent Boundaries and enhancements to Coder Tasks — address the struggles enterprises face as they adopt AI-assisted development without the infrastructure needed to manage the associated risks. Coder cited a recent Cisco Systems Inc. study that found that only 13% of global enterprises have a defined AI strategy, meaning that many teams rely on improvised setups that don’t scale.

“Bolting AI tools onto the old model, where code lives on local laptops, creates risk, cost and chaos,” Coder Chief Executive Rob Whiteley said in a statement. “This gets worse when you add AI agents, which are simply impossible to run concurrently on laptops.”

Coder is “kind of profiting on the fact that most folks probably don’t have an ideal AI strategy,” said Ben Potter, Coder’s vice president of product. “Developers are starting to speak and see what works, and this lets platform engineers become a little more strategic rather than reactive,” he said.

Agent equality

The company argues that organizations need a governed foundation capable of treating both human developers and AI agents as first-class participants in the development workflow. While acknowledging that several vendors offer broad, organization-wide controls, Potter said those toolsets lack visibility into one of the most sensitive layers: developer workspaces.

The result, he said, is that many enterprises can’t answer even basic questions about AI development, such as how much it costs, what tools developers are using and how to introduce coding assistants at scale.

“Coder, first and foremost, is in the developer environment,” Potter said. “We created a lot of these governance and controls because other vendors simply aren’t, at least around AI coding tools.”

AI Bridge consolidates authentication, access and usage data across all AI model providers used in development, replacing ad hoc proxies that lack centralized observability. It provides platform engineering teams with a single view of prompt logs, token consumption and model activity.

Agent Boundaries enforces policy-driven controls that restrict what AI agents can reach. The feature functions as an agent-specific firewall, Potter said. Organizations can define explicit lists for allowed network destinations, tools and internal systems. “If it a tool doesn’t have access to specific sources, risk is reduced significantly,” he said.

Coder Tasks expands the platform’s automation layer to support both human-initiated and agent-initiated jobs, enabling long-running, low-interaction workflows such as documentation, test authoring and code review.

Beyond laptops

A key part of the company’s pitch is that AI-augmented development shouldn’t rely on laptops as the primary execution environment. Coder’s main workspace platform allows enterprises to standardize developer runtime environments on self-hosted infrastructure. The new features extend this to AI agents, a shift Potter said customers requested.

“We’re trying to get development off of laptops,” he said, asserting that local machines can’t safely or consistently support the concurrent execution of human and agent workflows or support the controls enterprises require.

Coder relies on OpenID Connect and supports established single-sign-on systems to integrate with enterprise access controls. Developer workspaces are defined using Terraform templates, which embed organizational policies and required tooling.

Coder’s platform currently supports roughly 3,000 concurrent workspaces, equivalent to about 10,000 developers at a typical enterprise, and Potter said the company expects that scale to increase.

Pricing for the new capabilities has not been finalized, though Potter said enterprise customers evaluating large-scale agent deployments may eventually see per-agent pricing, while others will continue to pay per developer seat.

Coder plans to position its platform as the foundation on which both developers and AI agents can work side by side. “We’re not claiming to be a cohesive solution for everything,” Potter said. “We’re adding the layers enterprises need, so they can roll out AI tools with confidence.”

Founded in 2017, the company has raised more than $85 million in four funding rounds, according to Tracxn.

Image: SiliconANGLE/Ideogram