agentos.guide › blog

AI SEO Training: Where To Get It, And The 7 Modules Inside My Agent OS Pipeline

By Julian Goldie · 11 October 2026 · agentos.guide

AI SEO training is available free from me today, and the quickest start is my free AI SEO course, which currently shows 262 lessons across 22 modules.

I'm Julian Goldie, and I run SEO across my own sites with AI agents inside an Agent Operating System.

This page tells you where to get the training, and then it teaches the seven modules the way my Agent OS actually runs them.

You can follow along by hand with Search Console and any AI assistant.

Where to get AI SEO training

Here is every place I teach AI SEO, with the free options first.

If you're new, start with the free course and the AI Money Lab.

If you want the pipeline on this page installed and ready, it lives inside the AI Profit Boardroom.

The rest of this page is the training itself, taught through the Agent OS.

What AI SEO training covers, mapped onto the Agent OS

AI SEO training has seven modules, and each one maps onto a stage of the pipeline my agents run.

The Agent OS SEO tab runs a loop I call Read, Mine, Make and Ship, and then the results loop back to Read.

You can run every module by hand with Search Console and an AI assistant.

The Agent OS simply puts all seven behind one tab.

Quick start: your first loop in the Agent OS

This is the fastest route through the whole curriculum.

• First, add your site to Google Search Console and verify it, which is free.
• Second, open the SEO tab in the Agent OS and connect Search Console once, with read-only access.
• Third, press Refresh on Search OS to see your clicks, impressions and positions.
• Fourth, open Research, pick your site and press Run research.
• Fifth, take the first striking-distance keyword from the list.
• Sixth, open Generate, type the keyword and pick a transcript from one of your own videos.
• Seventh, press Deploy, and check History for the live URLs.
• Eighth, run research again next week and take the next keyword.

One loop a week is enough to start.

In my own recorded run, one keyword went from transcript to live, linked, indexed and tracked in about 25 minutes.

Module 1: keyword research with AI and Search Console

Search Console is the only tool that knows what Google already thinks of your site.

The Research tab reads it live and scores every query into three groups.

Striking-distance keywords sit between positions 5 and 20, where one good post can move you to page one.

Click-through leaks are keywords where you are already on page one and almost nobody clicks.

Content gaps are searches where you get impressions and have no dedicated page.

On 22 September 2026 the tab showed 78 striking-distance keywords and 23 leaks across my sites.

One line from that run shows the value.

The keyword "hermes workspace" had about 37,000 impressions across two sites and only 307 clicks.

Both pages were already on page one, so the verdict was to fix the title and not to write a new post.

In an earlier live run, the keyword skill scored 184 real queries in one pass and flagged a page at position 6.3 with no clicks.

If your Search Console is empty, you start with Google autosuggest instead.

You type a seed, add one letter at a time and save every suggestion, because each one is a real search.

You also save the People Also Ask questions, because they become your FAQ later.

🔥 Want the exact AI SEO training setup I run inside my Agent OS?

Inside the AI Profit Boardroom you get the installable Agent OS, step-by-step video tutorials, four coaching calls a week and 3,400+ members building real automations.

→ Get access here

Module 2: matching search intent

The skill file makes the agent classify the keyword before it writes anything.

The rule behind every row is that the page answers the core question on the first screen.

This page follows it, because the first line tells you where to get AI SEO training.

I also check page one by hand for sixty seconds per keyword.

A forum thread, a passing mention, an old page or a thin page tells me the door is open.

A page one full of big brands built for that exact keyword tells me to skip it.

Module 3: writing with AI the right way

The rule at the top of my skill file is that no transcript means no article.

I record my screen while I do real work, and each recording becomes a text file the agent can read.

The pipeline had 82 real transcripts in it when I recorded the spam update guide.

The agent may not invent a feature, a number or an example that is not in the source.

That is what makes the content different from the average of the internet.

Then the 13-step skill runs.

It covers the source of truth, the intent, the title and meta description, the draft, the keyword spread, the links, the readability and the fact check.

In one recorded test I planted four problems in a draft, and the check caught all four in about ninety seconds.

One was a line that said "300 clicks" when the case study said 278.

The last step is the one I like most, because the skill writes my corrections back into its own rules.

That means the skill I run in month three is sharper than the one I started with.

A human still reads the page at the end and asks whether it is something to be proud of.

Module 4: AEO and AI search citations

AEO means getting your pages quoted in AI answers.

Google's documentation says there are no additional requirements to appear in AI Overviews or AI Mode, and no special optimisations are necessary.

So the pipeline does not chase tricks.

It builds every page with the traits AI answers tend to lift.

• The page answers the question in its opening lines.
• The headings are phrased the way people ask.
• The FAQ answers are direct, self-contained and under about fifty words.
• The page carries original information from real work.
• The page is fresh, because it is published soon after the topic appears.

The agent also adds Article and FAQ schema to every post.

I'll be straight that Google's generative AI guide says structured data is not required for generative AI search.

I keep it because it takes the agent seconds and it labels the page clearly for machines.

My king of AEO page goes deeper on this module.

Module 5: internal and network linking

Before writing, the agent scans every existing post on the site.

Each new article links to two to four older ones, which shows depth on one topic.

After the article goes live, the agent edits three to five older related posts.

It adds a small "latest updates" block to each one that links to the new article.

So the new page gets links on day one, and the old pages get a freshness update.

I run several sites, so some of those blocks sit on my other properties.

Each post also carries a short network block that links to the same keyword on my sister sites.

My agents publish additively, which means they only add pages and never delete or overwrite a live one.

These are links between pages I own, and I stay away from link schemes because they break Google's spam policies.

🔥 Want the exact AI SEO training setup I run inside my Agent OS?

Inside the AI Profit Boardroom you get the installable Agent OS, step-by-step video tutorials, four coaching calls a week and 3,400+ members building real automations.

→ Get access here

Module 6: indexing

Google can take weeks to find a new page on a small site.

So the Deploy step ends by submitting every new URL for indexing.

The fastest result I've recorded is a brand new page on page one of Google in 14 hours.

That is my best case, and most pages take longer.

If you're not using the Agent OS, use the URL Inspection tool in Search Console and request indexing by hand.

Module 7: measuring and closing the loop

The Search OS panel shows the whole footprint on one screen.

It has a traffic chart for the last 28 days, a field of every keyword you already rank for and a queue naming the next three moves.

Each week I look for three things.

• I look for gaps, which are pages just off page one that need a fresh section and new internal links.
• I look for leaks, which are pages with rising impressions and flat clicks that usually need a better title.
• I look for winners, because every keyword that brings traffic becomes new seeds.

Every post you ship creates new impressions and new queries.

The next research run finds them, so the list of things to write never runs out.

That is why I say your rankings write the next brief.

The receipts behind this AI SEO training

Every number here is from my own Search Console and is shown on screen in a guide on this site.

I'll give you the honest timeline too.

New sites often need three to six months before the graph moves.

One of my sites stayed flat for about seven months before it climbed.

None of these numbers is a promise for your site.

Google's spam policies on scaled AI content

Any AI SEO training that skips this part is leaving you exposed.

Google's spam policies define scaled content abuse as generating many pages for the primary purpose of manipulating search rankings and not helping users.

One of Google's listed examples is using generative AI tools to generate many pages without adding value for users.

Google says that applies no matter how the content is created.

Google's AI content guidance says its focus is on the quality of content rather than how content is produced.

So the Agent OS is built to stay on the right side of that line.

Every post starts from a real transcript, passes 13 checks and gets read by a person.

Google ran spam updates in March, June and August of 2026, and I covered the August one in my spam update guide.

Other reputable AI SEO training options

I teach this for a living, so I'd rather you compare.

These providers are well known, and I've only listed facts shown on their own pages.

Google's own documents are the best source for the rules.

The academies are good for structured lessons and certificates.

My training is built around running the work with agents every day.

FAQ

Where can I get AI SEO training for free?

Julian Goldie's free AI SEO course currently shows 262 lessons across 22 modules. His AI Money Lab community, his YouTube channel and the Claude AI SEO course on agentos.guide are also free. Google Search Central, Ahrefs Academy and HubSpot Academy publish free training too.

What does AI SEO training cover?

It covers seven modules: keyword research with AI and Search Console, search intent, writing with AI, AEO and AI citations, internal and network linking, indexing and measuring.

How does the Agent OS run AI SEO?

The SEO tab reads Search Console, scores every keyword, writes posts from a real transcript using a 13-step skill, deploys them and submits the URLs for indexing. The new rankings then feed the next research run.

Does Google penalise AI content?

Google says its focus is on the quality of content rather than how it is produced. Its spam policies do penalise scaled content abuse, which means many pages generated to manipulate rankings without helping users.

Do I need the Agent OS to do AI SEO?

No. You can run all seven modules by hand with Google Search Console and any AI assistant. The Agent OS puts the same steps behind one tab.

How long does AI SEO take to work?

There is no guaranteed timeline. New sites often need three to six months before traffic moves, although Julian's fastest single page reached page one of Google in 14 hours.

Final word on AI SEO training

Take the free course, run one loop of the seven modules this week, and let your own Search Console guide the rest of your AI SEO training.

Related guides on agentos.guide

→ The Full Claude AI SEO Course

→ Claude AI SEO: Rank #1 Fast

→ The Infinite Ranking Engine

→ Julian Goldie SEO

📺 Video notes + links to the tools 👉

🎥 Learn how I make these videos 👉

🆓 Get a FREE AI Course + Community + 1,000 AI Agents 👉

← all guides on agentos.guide