> Markdown version of [/videos/2084-your-infrastructure-is-not-a-playground-ai-agents-for-infra-done-right?t=648](https://www.wearedevelopers.com/videos/2084-your-infrastructure-is-not-a-playground-ai-agents-for-infra-done-right?t=648). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Your Infrastructure Is Not a Playground: AI Agents for Infra Done Right

Stop letting manual errors trigger massive network outages. Learn how Agentic Ops safely pairs AI reasoning with deterministic tools and strict human-in-the-loop validation.

- **Speakers:** [Alfonso Sandoval Rosas](https://www.wearedevelopers.com/@alfonso-sandoval-rosas)
- **Event:** World Congress 2026 Europe - Virtual Stage
- **Published:** July 2, 2026
- **Duration:** 47:39
- **URL:** https://www.wearedevelopers.com/videos/2084-your-infrastructure-is-not-a-playground-ai-agents-for-infra-done-right

## Summary

Modern infrastructure operations frequently suffer from fragility, where simple manual errors can cascade into massive outages, forcing engineers to spend more than half their time on repetitive housekeeping tasks. Despite the evolution of infrastructure as code and API-first designs, engineers often remain hesitant to hand full control of legacy systems to automated scripts out of fear they might go rogue. The emergence of "Agentic Ops" addresses this challenge by employing AI agents capable of reasoning through high-level operational intents and safely breaking them down into sequenced, actionable tasks. Rather than replacing human oversight, these systems act like junior engineers under strict supervision.

At the core of this transition is the Model Context Protocol (MCP), functionally serving as a "USB-C port for AI." MCP streamlines how LLMs interact with real-world infrastructure by dynamically feeding them tool contexts and permissions without overloading prompt windows. By wrapping established network libraries—such as Cisco's PyATS—into dedicated MCP servers, organizations bind the probabilistic intelligence of large language models to deterministic, reliable tool sets. This separation ensures that the AI can plan heavily but can only execute actions permitted by strictly defined, read-only or authorized boundaries.

To build operational trust, this architecture heavily relies on layered guardrails and human-in-the-loop validation. Using low-code deployment platforms like n8n alongside collaboration tools like Slack, engineers can construct robust ChatOps workflows for continuous network management. Routine queries, compliance auditing against security guidelines, and automated GitHub issue tracking run invisibly in the background. However, when an agent intends to commit a configuration change, it executes a dry run, checks for issues like overlapping IP blocks, and pauses execution. A human operator then reviews the structured payload card in their chat interface and firmly clicks to approve or reject, ensuring that infrastructure remains a tightly controlled environment rather than a playground.

**Keywords:** agentic ops, infrastructure automation, MCP framework, model context protocol, human-in-the-loop validation, chatops workflows, infrastructure guardrails, deterministic tool execution, low-code deployment platforms, network compliance auditing, automated issue tracking, LLM context integration, legacy CLI operations, device inventory management

## Chapters

1. **The fragility and challenges of modern infrastructure management** (00:00) — Why manual housekeeping operations put critical network data centers at risk.
1. **The evolution of infrastructure management methodologies** (03:07) — How network operations shifted from human-centric CLI commands to intent-based AI agents.
1. **Overcoming barriers to infrastructure automation and AI** (07:38) — Tackling legacy devices, complex brownfield environments, and technical skills gaps.
1. **Applying large language models to infrastructure tasks** (10:48) — Using pre-trained LLMs to reason about system intents and execute low-code sequential steps.
1. **Contextualizing LLMs with the model context protocol** (12:57) — Integrating live systems using the universal MCP standard without extensive API training.
1. **Prototyping deterministic agents with n8n and PyATS** (18:21) — Leveraging visual node-based workflows and Cisco's open-source library for predictable interactions.
1. **Creating a basic network inventory agent in n8n** (22:06) — Connecting an LLM memory module and an HTTP MCP client to query live infrastructure data.
1. **Designing safe chat-ops workflows with human validation** (25:37) — Enforcing infrastructure safety by blending probabilistic LLM planning with strict deterministic Slack approvals.
1. **Automating network auditing and issue tracking with agents** (39:31) — Scheduling agents to scan configuration compliance files, output markdown reports, and generate GitHub tickets.
1. **Best practices for deploying safe infrastructure AI agents** (45:18) — Structuring simple workflows with strict persona boundaries to treat AI as a junior engineering assistant.

## Related Moments

- [Shifting security models from passive chatbots to active agents](https://www.wearedevelopers.com/videos/2093-from-shadow-ai-to-secure-intelligence-safe-ai-usage-in-the-enterprise) (from "From Shadow AI to Secure Intelligence: Safe AI Usage in the Enterprise")
- [Key learnings and infrastructure challenges with agentic automation](https://www.wearedevelopers.com/videos/100035-developers-become-orchestrators-from-human-in-the-loop-to-spec-in-the-loop) (from "Developers become Orchestrators: From Human-in-the-Loop to Spec-in-the-Loop")
- [Designing an agent workflow for infrastructure as code](https://www.wearedevelopers.com/videos/1533-infrastructure-as-prompts-creating-azure-infrastructure-with-ai-agents) (from "Infrastructure as Prompts: Creating Azure Infrastructure with AI Agents")
- [Hard science and infrastructure enabled by artificial intelligence](https://www.wearedevelopers.com/videos/100049-beyond-the-wrapper-technical-bets-that-vcs-back) (from "Beyond the Wrapper: Technical Bets That VCs Back")
- [Preserving engineering fundamentals in agentic development](https://www.wearedevelopers.com/videos/1897-agents-version-control-and-bunnies-daniel-siegl-david-payr) (from "Agents, Version Control and Bunnies - Daniel Siegl & David Payr")
- [Rethinking team structures around AI agent capabilities](https://www.wearedevelopers.com/videos/1539-agentic-devops-how-ai-powered-automation-transforms-software-delivery-on-github-and-azure) (from "Agentic DevOps: How AI-Powered Automation Transforms Software Delivery on GitHub and Azure")

## Related Articles

- [MLOps – What’s the deal behind it?](https://www.wearedevelopers.com/magazine/125-mlops-what-s-the-deal-behind-it)
- [MLOps And AI Driven Development](https://www.wearedevelopers.com/magazine/82-mlops-and-ai-driven-development)
- [What is Agentic Programming and Why Should Developers Care?](https://www.wearedevelopers.com/magazine/625-what-is-agentic-programming-and-why-should-developers-care)
- [MLops – Deploying, Maintaining And Evolving Machine Learning Models in Production](https://www.wearedevelopers.com/magazine/115-mlops-deploying-maintaining-and-evolving-machine-learning-models-in-production)

## Related Jobs

- [Senior Engineer, Infrastructure Platform](https://www.wearedevelopers.com/jobs/ext/328836-senior-engineer-infrastructure-platform) at **Intercom, Inc.**
- [AI Operations Manager (all genders)](https://www.wearedevelopers.com/jobs/48263-ai-operations-manager-all-genders) at **envelio**
- [AI Full Stack Engineer](https://www.wearedevelopers.com/jobs/ext/1354435-ai-full-stack-engineer) at **Almedia**
- [Senior Backend Developer — AI: MCP & Agent Engine](https://www.wearedevelopers.com/jobs/48297-senior-backend-developer-ai-mcp-agent-engine) at **basebox GmbH**
- [Senior AI Agent Software Engineer (Go, Python) (m/f/x)](https://www.wearedevelopers.com/jobs/48277-senior-ai-agent-software-engineer-go-python-m-f-x) at **Dynatrace**
- [Principal Field Architect - AI Agents](https://www.wearedevelopers.com/jobs/ext/1442858-principal-field-architect-ai-agents) at **Twilio**