OpsMill Infrahub: Infrastructure-as-Data for AI Agents

Infrastructure management is shifting from manually written scripts to autonomous agentic workflows. To support this transition, OpsMill today announced a $14 million Series A funding round to scale Infrahub, its pioneering "Infrastructure-as-Data" platform.

The CMDB for the Agentic Era

Traditional Configuration Management Databases (CMDBs) were built for humans to query. In contrast, Infrahub is designed as a graph-based source of truth for AI agents. By providing a version-controlled, schema-first representation of infrastructure, it allows agents to "reason" about dependencies before making changes.

Instead of an agent guessing which subnet is available, it queries Infrahub's graph. Because the data is versioned (similar to Git but for live infrastructure state), agents can propose changes in "branches," which are then validated by automated CI pipelines before being merged into the production state.

Infrastructure-as-Data vs. Infrastructure-as-Code

While Infrastructure-as-Code (IaC) like Terraform has been the standard, it often suffers from "state drift" and lacks a global relationship map. OpsMill's approach of Infrastructure-as-Data treats the desired state as a living graph. This makes it significantly easier for LLMs to navigate and manipulate without the brittle syntax errors common in generated YAML or HCL.

The funding round, led by top-tier venture firms, underscores the industry's belief that agent-native infrastructure is the next frontier of platform engineering.