Tracks Courses Interactive Roadmap Community Blog About Start AI course →
Free & open-source learning platform

Ops → AI

Every DevOps, SRE, and Platform engineer is asking the same question: where do I start with AI? This is the answer. Free courses, real projects, real code in Go. No fluff, no paywall.

Start learning → See the tracks
scroll
KubernetesAI AgentsLLM FoundationsRAG SystemsDockerGitHub ActionsTerraformMCP ProtocolPrometheusGo ProgrammingArgoCDOpenShiftVector DatabasesSRE PracticesAIOpsGrafanaLangChainGo KubernetesAI AgentsLLM FoundationsRAG SystemsDockerGitHub ActionsTerraformMCP ProtocolPrometheusGo ProgrammingArgoCDOpenShiftVector DatabasesSRE PracticesAIOpsGrafanaLangChainGo
Learning in public

What's new today

Bite-sized concepts I'm learning as I build this platform. New things, fresh ideas, real discoveries — scroll through.

⏳
Loading…
Share something new
💡🔁🧠⚡🚀🔧📦🤖🔑
01 — THE SHIFT

The ground is moving under Ops

AI isn't coming to operations. It's already here. Incident response, log analysis, runbook automation — all being reshaped. The engineers who adapt will lead. The ones who wait will be left explaining why they didn't.

02 — THE GAP

But the learning path doesn't exist

Every AI course assumes you're a Python data scientist. None of them speak to the engineer who lives in Kubernetes, Go, and bash — who wants to use AI in production, not in a notebook.

03 — THE BRIDGE

So we built the bridge

From your first container to your first AI agent. A clear, efficient path that respects what you already know and shows you exactly what's new. This is ops → ai.

0
Courses live
0
Planned courses
0
Free, forever
0
To get started
Two tracks, one destination

Pick your starting point

Whether you need to strengthen your Ops fundamentals or dive straight into AI — there's a path for you. They converge where the real opportunity is: AI-powered operations.

⚙️ Ops Core Track

The technologies every Ops engineer must know

The foundational stack that 82% of production environments run. Master these and you can work anywhere.

  • Linux & Networking2 wks
  • Docker & Containers1.5 wks
  • Kubernetes Fundamentals3 wks
  • OpenShift2 wks
  • CI/CD — GitHub Actions1.5 wks
  • Terraform / OpenTofu2 wks
  • Observability — Prometheus, Grafana, Loki2 wks
  • SRE Practices & Incident Management1.5 wks
  • GitOps — ArgoCD1 wk
Browse Ops courses →
🧠 AI Bridge Track

How to bring AI into your Ops work

Already know Ops? This is the fast track. Learn AI from an engineer's perspective, build real tools in Go.

  • LLM Foundations & Prompt Engineering1 wk
  • RAG & Vector Databases1.5 wks
  • MCP — Model Context Protocol1 wk
  • Building AI Agents in Go2 wks
  • AIOps — k8sgpt, kagent, kubectl-ai1.5 wks
Start the AI Bridge →
Courses

What's ready, what's next

Each course is project-based and built for engineers who ship — not students in a classroom.

🤖Live
AI for DevOps Engineers — Go Edition
LLMs, RAG, AI Agents, MCP, and AIOps — then build InfraBot, a real AI agent in Go that reads your runbooks and checks service health.
GoLangChainGoRAGMCPOllama
8 weeks · 4 theory + 8 project phasesStart →
🚀Live
OpenShift — Beginner to Advanced
Complete guide to Red Hat OpenShift: architecture, oc CLI, deployments, RBAC, Operators, CI/CD with GitOps, and cluster administration.
OpenShiftKubernetesRBACOperators
17 chapters · Beginner to AdvancedRead →
☸️Coming soon
Kubernetes Deep Dive
Production Kubernetes for platform teams — networking, storage, RBAC, multi-tenancy, Helm, and troubleshooting.
KubernetesHelmNetworking
~3 weeksComing soon
🔄Coming soon
CI/CD with GitHub Actions
Build, test, scan, and deploy — pipeline-as-code from scratch. The CI tool used by 62% of developers.
GitHub ActionsCI/CD
~1.5 weeksComing soon
📊Coming soon
Observability — Prometheus, Grafana & Loki
Metrics, logs, and traces. Dashboards, PromQL, alerts, and incident troubleshooting with real data.
PrometheusGrafanaOpenTelemetry
~2 weeksComing soon
🏗️Coming soon
Terraform / OpenTofu
Infrastructure as Code — HCL, state, providers, modules, and the Terraform vs OpenTofu licensing landscape.
TerraformOpenTofuIaC
~2 weeksComing soon
Not sure where to begin?

Find your starting point

Answer 3 quick questions and we'll point you to the right place.

🎯
Your recommended start
Learning path

The most efficient order

Don't wander. Follow this path. Each step builds on the last — designed so you can complete it part-time without burning out. Open interactive roadmap →

1
Start here
OpenShift or Kubernetes fundamentals
Understand containers, pods, and the platform. If your company uses OpenShift, start there. Otherwise, learn vanilla Kubernetes first.
2
AI Week 1–2
LLM foundations & RAG
Understand what LLMs are, how tokens work, and how to make AI answer from YOUR documentation using RAG. The most immediately useful AI skill for Ops.
3
AI Week 3–4
MCP, Agents & AIOps tools
Go from "AI answers questions" to "AI takes actions." Learn MCP, build an agent in Go, and discover tools like k8sgpt you can use at work today.
4
Ops deepening
CI/CD, Observability, IaC & SRE
Round out your Ops stack: GitHub Actions, Prometheus + Grafana, Terraform, ArgoCD, and SRE practices. Each course is self-contained.
5
Build & ship
InfraBot — your portfolio AI project
Build a working AI agent in Go that searches runbooks, checks services, and creates incident reports. Put it on GitHub. Write about it on LinkedIn.
Why this exists

The problem is real

I faced this problem. I saw others facing it. So I built this.

😤
The learning landscape is a mess
AI courses are built for ML researchers or Python devs. Nothing exists for Ops engineers who live in Go and bash and want to use AI at work — not in a Jupyter notebook.
🎯
You already have the hardest skill
You understand systems, reliability, and production. You know what an incident is. That context makes you incredibly well-positioned for AI — if someone shows you how.
🔓
Free, forever, no catches
No course fees, no subscriptions, no paywalls, no email gates. If these guides help you, share them with someone else who needs them.
⚡
Built with AI, curated by an engineer
Built using Claude AI and curated by a working engineer in DevOps and SRE. Real tools, real code, real concepts — not repackaged generic tutorials.
Who's behind this

Hi, I'm Rahul

I'm a Specialist DevOps engineer at Amadeus — four and a half years, and three levels in that time. Every one of those promotions came from the same habit: I notice where engineering teams are quietly drowning, and I go build the thing that gets them out.

Nobody usually asks. I just keep watching the same fifteen minutes get wasted, the same dashboard get checked by hand — and eventually I build something so it stops happening.

in LinkedIn ⌥ GitHub
RA
Rahul Agrawal
Specialist DevOps Engineer
NowAmadeus · near Munich, Germany
SinceOct 2021 · 3 promotions
BuildsGo · Kubernetes · AI agents
RHCSA AZ-900 M.Sc. Göttingen
84K+
Incidents handled by a platform I built and own, end to end
7,000hrs
Given back to engineers who were doing it by hand
3levels
Promotions in four and a half years, each one from building something unasked
15→1
Minutes to first action on an incident, down from fifteen

What I've built, and why

01
The incident platform
The problem Our incident team lost fifteen minutes on every single incident — hopping between ten tools before anyone even started fixing anything.

Nobody asked me to solve it. I just kept watching it happen. So I built a platform in Go, as the sole engineer, that automates that entire first step — including routing each incident to the right team through a rules engine. Backed by 100+ unit tests.

84,000 incidents. 7,000 hours given back. But the number I actually care about isn't incidents — it's trust.

GoAutomationRules engine100+ tests
02
An AI agent for engineering operations
The problem Engineers juggling Confluence pages, ServiceNow rosters, and dozens of Grafana dashboards — every answer buried somewhere different.

Now a lot of people don't dig through any of it — they just ask the agent. It's an MCP-based service in Go, running in production, that watches dashboards, detects anomalies, and answers questions about our operation in seconds.

MCPAzure OpenAIGoGrafana
🌐
And then there's this site

That builder instinct doesn't switch off on Friday. A while back I realised I'd used the web my whole life without ever building a full site from scratch — so I spent a weekend doing it hands-on. Auth, database, hosting, domain, all of it.

That became opstoai.com. Now it's just where I drop everything I learn about AI, so other people can pick it up faster than I did.

Our time and energy are limited. If we spend them on boring, repetitive work, we lose our motivation.

I don't care about being the fastest engineer in the room. I'd much rather build the things that make everyone around me faster.

If you're building something similar, or this site helped you, I'd genuinely like to hear about it — reach out on LinkedIn.

Ready to go
from Ops
to AI?

Start with the AI for DevOps course. It's free, runs locally with Ollama — no API keys, no costs. Just learning.

Start AI for DevOps → Or start with OpenShift
📬

Get notified when new courses drop

No spam, no weekly emails. Just a heads-up when a new course or major guide goes live. Unsubscribe anytime.

Built in public. Free forever.