AI won't take your IT job. The dev who uses it will.
Developers, sysadmins and support engineers who work with AI ship faster, break less, and clear tickets before lunch. In 2026 that's not a nice-to-have — it's the baseline hiring managers screen for. This track shows you exactly what to learn, and how to prove it.
Volumes to read.
Four volumes. MCP and Agents are the two most candidates can't do yet — that's the point.
AI Foundations
Start from zero: what AI is and how to ask it well.
MCP Mastery
Plug real files, tools and data into AI the standard way.
AI Agents
Build an AI that acts on a schedule, not just answers.
JARVIS
Your own AI second brain, wired to your real work.
The 2026 job posting already assumes you use AI.
Read any current IT job description and you'll see it: "experience with AI-assisted development," "familiarity with LLM tooling," "automation mindset." A few years ago that was a bonus line. Now it's the filter. Teams aren't hiring people to compete with AI — they're hiring people who get more done because they wield it well.
"AI-assisted" is in the spec
Copilot, Claude and Cursor are standard-issue on most engineering teams. Showing up not knowing how to drive them now reads the way "never used Git" did a decade ago. Employers expect fluency, not curiosity.
Output per engineer went up
A dev who pairs with AI closes more tickets, ships more features, and unblocks teammates faster. Managers see it in the throughput. The candidate who visibly does this wins the offer.
Experience gap, levelled
No 10 years on the job? AI closes part of that gap — you can read unfamiliar codebases, debug faster, and explain systems you didn't write. Show that in an interview and juniors suddenly look senior.
The AI skills that get you hired in IT.
Not "prompting tips." These are concrete, on-the-job capabilities a hiring manager can picture you doing on day one. Learn a handful of these well and you interview differently.
AI pair-programming that actually ships
Use Copilot and Claude as a pair to scaffold features, write tests, and get from "empty file" to "working PR" in a fraction of the time — then read the output critically instead of pasting it blind. The skill isn't generating code; it's steering it and knowing when it's wrong.
Debug & review at 3x speed
Paste a stack trace and get to root cause faster. Have AI review your diff for bugs, edge cases and security holes before a human ever sees it, and explain a gnarly legacy function line by line so you can safely change it. You look careful and fast at the same time.
Write & explain scripts on demand
Bash, PowerShell, Python, a fiddly regex, a Dockerfile, a CI step — describe what you need and get a working draft, then have it explain every line so you actually learn it. Perfect for the sysadmin and support work where you hit a new tool every week.
Automate ops & tickets with agents
Wire AI into n8n or an agent flow to triage incoming tickets, draft the first response, tag and route by priority, and run routine ops tasks on a schedule. This is the "automation mindset" line in the JD, made real — the thing that quietly saves your team hours.
Connect internal tools with MCP
Use the Model Context Protocol to plug AI into your real systems — the ticketing tool, the database, the internal wiki, the deploy scripts — so it can answer with live context instead of guesswork. This is a genuinely rare, in-demand skill most candidates can't do yet.
Turn docs into a searchable assistant
Point AI at your runbooks, READMEs and Confluence and turn a wall of scattered docs into something a teammate can just ask: "how do we roll back prod?" "where's the staging config?" Onboarding drops from days to minutes — and you're the one who built it.
A route through the library, built for IT.
You don't need all nine volumes to start interviewing better. Do these four in order — start free, and each one builds directly on the last toward a capstone you can show off.
- AI Foundations Free
How the models actually work and how to talk to them so the output is worth keeping. Every skill on this page stands on this one. Start here — it's free.
- MCP Mastery Core
Connect AI to real tools — databases, ticketing, internal docs, your own scripts — with the Model Context Protocol. This is the skill that makes an AI useful inside a real IT stack, and the one most candidates can't do.
- AI Agents Build
Turn one-off prompts into automations that run on their own — ticket triage, ops jobs, scheduled reports. This is the "automation mindset" employers keep asking for, built with your own hands.
- JARVIS — the capstone Capstone
Tie it all together into an AI second brain wired to real context — the piece that makes "knows AI" concrete on your résumé and gives you something live to demo in an interview. End here; it's what people remember.
See the full catalogue and every volume on the courses page →
One project that makes an interviewer sit up.
Talk is cheap and everyone claims they "use AI." Build this instead — a real, demoable tool that shows every skill on this page working together. It's the single best thing you can bring to an IT interview in 2026.
The internal "team assistant" that answers from your own docs
A chat assistant your team can ask "how do we deploy service X?" — and it answers from your team's real runbooks, READMEs and wiki, not the open internet. Onboarding a new hire goes from a week of asking around to a single conversation.
What you build
- Ingest the docs — point it at your READMEs, runbooks and wiki pages so it has the real answers.
- Wire it up with MCP — connect the assistant to those sources (and optionally the ticket tool) so it responds with live context.
- Add an agent layer — let it not just answer but do: open a ticket, flag a stale runbook, post the deploy steps.
- Ship a simple interface — a chat box or a chat-tool command anyone on the team can use.
Why it lands the offer
- It proves MCP + agents + RAG in one artifact — the exact skills the JD asks for.
- It solves a problem every IT team actually has, so managers instantly get the value.
- You can demo it live in the interview instead of describing a tutorial you followed.
- It's a natural first slice of your own JARVIS — reuse it forever, not just for the interview.
Become the hire who already knows AI.
Two next steps: get a CV that puts your AI skills up top where recruiters look — or build the JARVIS capstone that gives you something real to demo. Do both.