Former HashiCorp CEO Dave McJannet Launches Dome Systems to Govern AI Agents

Dave McJannet, the former CEO of HashiCorp, has co-founded Dome Systems to provide a governance layer for enterprise AI agents. The platform is designed to unblock the adoption of agentic AI by solving critical security and compliance hurdles that currently limit large-scale deployment.

Positioned as an "agent-native governance platform," Dome Systems serves as an operational control plane for the agentic enterprise. The system provides a unified compliance and control framework that operates across various agents, runtimes, and cloud environments. McJannet, who led HashiCorp through its IPO and subsequent acquisition by IBM, is leveraging his experience in infrastructure to address the unique requirements of autonomous AI systems.

The Governance Gap

Traditional enterprise software relies on deterministic workflows, where a specific input consistently triggers a predictable path and output. This model allows for rigid access-control systems that can precisely manage who or what can access specific data. However, AI agents are probabilistic, meaning they may employ different tools or sequences of actions to reach a result for the same query.

McJannet argues that this unpredictability breaks cloud-era governance. He compares the behavior of agentic apps to that of Waymo, noting that their probabilistic nature makes traditional deterministic access controls insufficient. Because these "reasoning systems" behave more like employees than static software, they create a governance gap that existing security tools are not equipped to handle.

Scaling Agentic AI

This shift in how software operates represents what McJannet frames as the next major platform transition following the cloud era. For large organizations, the inability to guarantee compliance and security for autonomous agents means that deployment is often restricted to low-risk pilots. Without a way to enforce safety rails, the risk of unpredictable agent behavior prevents the transition from experimental use to core business operations.

By creating a dedicated governance layer, Dome Systems intends to provide the necessary oversight to allow enterprises to deploy agentic AI at scale across their entire infrastructure. The goal is to move AI agents out of isolated sandboxes and into production environments where they can interact with enterprise data securely.

What's Next

As Dome Systems rolls out its operational control plane, the industry will be watching to see if a centralized governance layer can truly mitigate the risks of probabilistic AI. The success of the platform will likely depend on its ability to integrate seamlessly across diverse cloud providers and agent runtimes while maintaining the flexibility that makes agentic AI valuable in the first place.