IP Fabric launches MCP server for network operations

IP Fabric has launched a Model Context Protocol server for network operations, aimed at organisations adopting AIOps in regulated enterprise environments.

Deployed from the IP Fabric appliance user interface, the server connects AI tools to the company's network digital twin. Setup is opt-in by default and includes a prompt library designed to help teams apply AI to operational tasks.

The launch comes as businesses test broader use of AI in infrastructure management while facing concerns over data access, governance and the reliability of automated actions. IP Fabric is positioning the product around those issues, arguing that enterprise users need tighter control over how AI systems interact with network information.

The server gives AI systems access to a continuously updated model of network and cloud environments, including devices, connections and configurations. That model can be used for troubleshooting, compliance checks and assessing the likely effect of changes before they are introduced into live systems.

One feature IP Fabric highlighted is governed access to network data. In practice, organisations decide whether to enable the server and how AI tools can use the underlying information, a point likely to matter in sectors with strict regulatory requirements.

Another part of the launch is a set of pre-built prompts designed around common operational workflows. These are intended to give teams a starting point for using AI in network operations without having to build every interaction from scratch.

"Enterprises are racing toward network automation, only to question if they can trust AI to deliver an outcome that won't inadvertently damage digital business," said Pavel Bykov, Chief Executive Officer, IP Fabric.

"We built our MCP server to solve that problem at enterprise scale. By combining secure deployment, governed access and a library of proven prompts, we give teams a safe and practical way to securely and successfully bring AI into network operations," Bykov said.

Network Model

IP Fabric's platform is built around a digital twin of an organisation's network estate. Its software continuously discovers and models network state, then validates it against intended business and operational policies.

By exposing that information through the MCP server, users outside traditional network teams can query the environment in natural language. That could extend access to engineers who are less familiar with the platform, as well as non-technical users seeking visibility into network conditions.

IP Fabric also said the digital twin is vendor-neutral and can cover the full network estate, including cloud environments. According to the company, that makes it possible to test automated compliance checks, examine AI-generated decisions and run scenario modelling before implementing changes in production.

Use Cases

Among the use cases cited are compliance and configuration drift detection. The system can compare live network state against intended design and assess alignment with security and regulatory frameworks including CIS, NIST, ISO 27001, PCI-DSS, HIPAA, NIS2 and DORA.

IP Fabric also pointed to automated troubleshooting and root cause analysis, where AI agents could identify where network behaviour differs from expected intent. A third area is change planning, including projects such as SD-WAN migrations, mergers and acquisitions integration work, and cloud adoption programmes.

Those examples reflect a broader trend in IT operations, where suppliers are trying to move AI beyond general-purpose assistants toward systems tied to operational data and controls. In network management, the challenge has been to make AI useful without creating new security, compliance or reliability risks.

IP Fabric did not disclose pricing in the announcement. It also did not provide customer deployment figures for the new MCP server, though it said the wider platform is used by organisations including Red Hat, Major League Baseball and Air France.

The company said its core platform creates a unified view of devices, configurations, state and dependencies across complex environments within minutes, with automated checks designed to reveal gaps between intended and actual conditions.

For enterprises weighing AI in operations, the significance of products such as this will depend on whether they fit within existing governance models and deliver measurable reductions in manual work. IP Fabric's argument is that a controlled link between AI tools and a validated network model offers a safer route into that shift.

For now, its pitch is narrower: giving network teams and adjacent users a governed way to query infrastructure data, investigate problems and assess planned changes using AI grounded in a live operational view of the network.