Everyone has ChatGPT. Almost nobody uses more than a fraction of it.
This is the training that closes the gap — prompting that returns usable output, custom GPTs built around your actual work, and the judgment to know when to stop trusting it. Learn it yourself, or bring us in to train your team.
Whether you learn it yourself or we run it for your team, this is the ground it covers. Roughly in this order, because each piece depends on the one before it.
01
Prompting fundamentals
The structure behind a request that returns something usable. Context, constraints, examples, and format — the difference between a mediocre first draft and one you can actually send.
Framing context so the model has what it needs
Constraints that prevent generic output
Iterating instead of restarting
02
Custom GPTs
Stop re-explaining your business every session. Build assistants that already know your product, your tone, and your process.
Packaging your context once
Assistants for recurring work
Sharing them across a team
03
Research and analysis
Where ChatGPT saves the most time for most people — and where it will confidently hand you something wrong.
Competitor and market research
Summarizing long documents without losing the point
Checking output you cannot personally verify
04
Writing and communication
First drafts, rewrites, and tone shifts. The goal is not AI-sounding copy; it is getting to a draft you can edit rather than a blank page.
Drafts that sound like you
Rewriting for a different audience
Where to stop and write it yourself
05
Working with data
Spreadsheets, exports, and messy files. ChatGPT handles more of this than most people realize, and fails at some of it in ways worth knowing in advance.
Cleaning and reshaping exports
Analysis and charts from raw files
When to stop trusting the arithmetic
06
Where it breaks down
The most useful thing training gives you is judgment about when to close the tab and use something else.
Tasks ChatGPT is genuinely bad at
When to reach for Claude instead
What never goes into a public model
That's the ground. Cover it yourself, or have us run it for your team.
Start cheap and see if it clicks, or go straight to the full path. The Quickstart is included in the course, so starting small costs you nothing if you upgrade later.
Start here
The AI Agent Quickstart
$35$99
Launch price
My 15 working agents — email triage, call prep, daily briefing, contract review — packaged to install in Claude Code in minutes. Plus 12 months of community. No subscription, no course to finish.
Includes the AI Agent Quickstart + 12 months of community
The flagship. Master Claude Code, Claude Cowork, and ChatGPT Codex — and leave with 5+ working agents, your own Command Post, and the Certified AI Agent Manager credential. Next cohort starts August 12.
Training one person is nice. Training the team changes the numbers.
A ChatGPT workshop gets everyone to the same baseline in a day — same prompting framework, same shared library, same understanding of what not to paste into a public model. Onsite or virtual, built around your team's real work.
If you are using ChatGPT to reword emails, no — you can figure that out yourself. Training is worth it when you want the gap closed between casual use and the workflows that save real hours: structured prompting, custom GPTs built around your work, and knowing where the tool quietly gets things wrong.
Can I not just learn ChatGPT for free?
Yes, and plenty of people do. The free route costs you time and a lot of trial and error, and most people plateau at surface-level use without realizing it. Paid training buys you the sequence — what to learn in what order, and which habits to skip.
How long does it take to get good at ChatGPT?
A focused hour gets you past the beginner plateau. A few weeks of deliberate use makes it a genuine part of how you work. The people who never get there are usually not lacking time — they never changed how they prompt.
Do I need any technical background?
No. Nothing here assumes coding, and most people we train have never opened a terminal. If you can write a clear email, you have the skill this builds on.
ChatGPT or Claude — which should I learn?
Learn ChatGPT first; it is where most teams already are and the skills transfer. Claude tends to be stronger for long documents, careful reasoning, and actually building things via Claude Code. Most people who work with AI seriously end up using both for different jobs.
Can you train our whole team on ChatGPT?
Yes. Team workshops run half-day or full-day, onsite or virtual, and are built around your team’s real workflows rather than generic demos. That is a separate engagement from the self-paced courses — book a call and we will scope it.
What is included in the self-paced option?
The AI Agent Quickstart is 15 working agents packaged to install in minutes, plus 12 months of community access. If you want the full path — including live build sessions and the certification — the AI Agents Course includes the Quickstart.
Will this be out of date in six months?
The specific interface will change. The underlying skill — how to structure a request, when to trust the output, where to keep a human in the loop — has been stable for years and transfers across tools and model versions.
Is it safe to use ChatGPT with work data?
It depends entirely on your plan and settings, and this is where most companies get burned. Personal accounts and enterprise accounts behave very differently regarding data retention. Both the self-paced material and team workshops cover exactly what to check before connecting anything.
What is the guarantee?
Self-paced purchases carry a 14-day money-back guarantee. Team workshops carry a money-back guarantee on measurable productivity — details are agreed on the call before anything starts.