We Just Brought Generative AI to Network Configuration — and We’re Giving It Away for Free

Today marks a huge milestone for us at rConfig — and, honestly, for the network automation industry as a whole.

Live from Zabbix Summit 2025 in Riga, we’re proud to announce two industry-first AI capabilities for network configuration management:

• rConfig GenAI — in-app Generative AI designed to help engineers understand, troubleshoot, and optimise their configurations and network state.

• rConfig MCP AI — a Model Context Protocol integration that opens up a completely new developer experience for network automation.

Both are available right now, built directly into rConfig V8. And both are completely free.

No paywalls.
No “Pro+AI” license tiers.
No monthly tokens.

Just powerful, open AI tools that you control — on your infrastructure, with your data, using your own AI services.

Why We’re Doing This

The last few years have seen AI find its way into every corner of IT — but not always in ways that serve engineers. Most “AI for ops” tools are locked behind subscriptions, usage fees, or questionable data pipelines.

We’ve seen too many great ideas spoiled by bad incentives. AI shouldn’t mean giving up privacy or signing up for another recurring cost.

So, when we began building AI capabilities for rConfig, we set one non-negotiable rule: engineers should stay in control.

“AI should empower engineers, not bill them monthly or harvest their configs,” says Stephen Stack, CTO of rConfig.
“Our philosophy was simple — your data, your AI, your terms.”

That’s why rConfig’s new AI features are completely open and BYO-AI (Bring Your Own AI). Whether you use OpenAI, Anthropic, Gemini, or any other compatible service, rConfig connects securely to it — and nothing leaves your environment unless you choose it to.

No data is stored, logged, or shared by us. Everything runs with your authentication, on your schedule, in your own workflows.

This isn’t “AI as a Service.”
It’s AI as a Skill — built into your network toolkit.

Meet rConfig GenAI

Let’s start with the first half of this release: rConfig GenAI, our new built-in assistant for configuration and state analysis.

Imagine being able to ask questions like:

• “Why is this routing table changing every few minutes?”

• “What’s the difference between these two BGP configs?”

• “Show me any security policy drift over the last week.”

• “Explain this config section to a junior engineer.”

GenAI reads your actual configuration and historical state data, surfaces context, and responds in natural language. It’s the assistant that doesn’t just parse syntax — it understands network intent.

You can use it to identify compliance gaps, troubleshoot anomalies, or even teach new engineers why a config behaves a certain way. It turns static text into live, explorable knowledge.

“The power of this feature is how human it feels,” Stephen adds.
“You’re not trawling through logs or diff files. You’re having a conversation with your network.”

Under the hood, rConfig uses your chosen AI service through a secure connector — meaning none of your configuration data ever leaves your environment unless you explicitly permit it.

That’s a first in the NCM space.

Introducing MCP AI — For Builders, Developers, and Tinkerers

The second half of this announcement is for the developers, automation engineers, and integration specialists.

We’ve integrated Model Context Protocol (MCP) — the same emerging standard that’s reshaping how AI agents interact with tools and APIs.

rConfig’s MCP AI integration makes our platform discoverable and callable by AI coding tools like Claude Code, Codex, or Gemini Code Assist.

In simple terms, you can now connect an AI assistant directly to your rConfig environment and say things like:

• “Generate an API call to back up all Cisco devices in group X.”

• “Write a Python script that compares yesterday’s VLAN assignments.”

• “Show me how to use the compliance API to check SNMP configurations.”

The AI can respond with usable code, examples, or automation templates — all while respecting your local rConfig permissions and context.

For teams building integrations, extensions, or analytics, this is huge. It means developers can explore rConfig’s codebase, data models, and APIs conversationally. No need to dig through documentation or reverse engineer endpoints.

“Our open-source DNA made MCP a perfect fit,” Stephen explains.
“It bridges what’s in the code with what’s in your head. You can literally ask your AI to understand rConfig like a senior developer would.”

We see this becoming a cornerstone for partners, MSPs, and enterprise customers building their own automation on top of rConfig.

Built on Openness — and Built to Last

Both AI capabilities build on the foundation we laid with rConfig V8, launched earlier this autumn.

V8 introduced a complete modernization of our platform — from performance and search to testing and compliance. Over 6,000 automated tests, zero CVEs, and our fastest configuration engine ever.

The new AI features don’t replace that foundation — they elevate it.

They bring context, conversation, and intelligence to what was once a manual, log-heavy workflow.

And they do it without changing how you deploy, license, or secure rConfig.

No hidden tiers. No separate editions.
Just one version — rConfig for everyone.

“We’ve seen too many vendors slice their features into upsells,” says Stephen.
“That’s not who we are. If we believe a capability makes network engineers more effective, it belongs in the product — not behind a paywall.”

What This Means for Network Teams

Let’s break down a few real-world scenarios where rConfig’s new AI capabilities change the game:

1. Smarter Troubleshooting

Network instability often starts with subtle configuration drift. GenAI can now explain those differences in plain English, highlight risks, and even suggest where to look next — saving hours of manual diff analysis.

2. Guided Learning for Junior Engineers

Mentoring new staff is time-consuming. With GenAI, they can ask the “why” questions directly inside rConfig — and get accurate, contextual answers from real configurations. It’s like pairing every new hire with an experienced mentor.

3. Safer Change Control

Before a major update, teams can ask rConfig AI to analyze the proposed configuration against known best practices or compliance rules. It surfaces hidden risks — without needing a dedicated compliance engine or external parser.

4. Accelerated Development

With the MCP AI integration, dev teams can generate workflow scripts, test new API calls, and build internal tools using natural language. You can literally ask:

“Write me a PHP script that pulls compliance history for all devices in the ‘datacenter’ group.”
and get working code — instantly.

5. Secure Autonomy

All of this happens inside your infrastructure, authenticated with your AI provider, under your security controls. No external processing. No AI vendor lock-in.

That combination — power, privacy, and portability — is what makes this release unique.

A First for the NCM Industry

To our knowledge, no other Network Configuration Management tool — open source or commercial — offers in-app Generative AI or MCP AI support.

That’s not marketing bravado; it’s the reality of a market that’s been cautious, even conservative, about AI adoption in operational tools. And for good reason — security, data sensitivity, and change control are non-negotiable in networking.

rConfig bridges that gap safely.
It’s the first implementation where AI isn’t a risk — it’s a responsible companion.

“This is a line in the sand for us,” Stephen says.
“We’re not waiting for the big vendors to define what AI in networking should look like. We’re defining it ourselves — open, transparent, and engineer-first.”

The Road Ahead

These AI features are just the start. Over the next year, rConfig will expand its AI framework to support more models, context-aware insights, and automation templates for specific vendors and industries.

Our goal is to make network configuration and state management not just automated, but self-aware — where the system helps you understand why things happen, not just what happened.

At the same time, we’re staying true to what’s made rConfig stand out for over a decade: open-source values, security-first design, and pricing that actually makes sense.

**