ChatGPT training - OpenAI

ChatGPT course: get real work out of the tool you already have

The prompting method first, then the ChatGPT path, then the automations that run without you. Online, at your own pace.

Start now, 290 € 15 free lessons
9.4 / 10 from 497 reviews
82 lessons, 7 of them on ChatGPT
8h 40min of reading and exercises
chatgpt.com - a GPT being built
The ChatGPT builder: name, description, instructions, knowledge base

What you learn to drive

ChatGPT, screen by screen

Three places six months of use will not show you, and they are what separates chatting from serious work.

The ChatGPT builder: name, description, instructions, knowledge base
  • Create, or configureTwo ways to build a GPT: describe it in conversation, or fill the fields yourself. The lesson shows why the second gives you something you can actually correct.Lesson 3: the custom assistant, built in front of you.
  • InstructionsWhat the GPT does, how it behaves, what it must avoid. This is where the prompting method settles once, instead of being retyped in every conversation.Prompting module: the five parts, applied.
  • The knowledge baseYour files, loaded into the GPT. The lesson is blunt about what that means: conversations can reveal their contents, so you choose what goes in.Module on data and rules of use.
  • The previewYou try the GPT while you set it up, in the right-hand panel. The back and forth between instructions and result is the actual work.Lesson 3: set it, try it, fix it.

Real documents

What comes out of the course

Five documents produced during a session, in front of the room, from one company’s own material. Open them: this is the kind of thing you will be handing over.

A meeting preparation note

A meeting preparation note

A forty-two page sector report in, a one-page briefing note out, with the source of every figure.

42 pages read, a note in five minutes

Open the document ->
A board meeting

A board meeting

Scattered notes become eight slides and three decisions to take. One invented figure along the way, caught on review.

scattered notes, eight slides

Open the document ->
A monthly dashboard

A monthly dashboard

The accounting export and the quotes export, cross-checked into five figures that fit on one screen. No manual entry.

two raw exports, fifteen minutes

Open the document ->
A one-page website

A one-page website

A full page from a single request, then adjusted in three sentences. What the tool can do, and where it stops.

one request, three corrections

Open the document ->
A tender response

A tender response

Research on the buyer, more research in your own files, and a submission. Including a certification wrongly claimed, then corrected.

two searches, a full submission

Open the document ->

The method, applied to ChatGPT

The same request, written twice

What the course changes sits in this gap. Five parts to put into a request: context, role, action, format, tone.

What most people type

YouWrite me an email to follow up with a client.

Subject: Following up on our proposal. Hello, I hope this email finds you well. I am reaching out regarding the proposal we sent recently. Please do not hesitate to share your feedback. Looking forward to hearing from you.

The same one, once the method is learned

You

The ChatGPT path

The 7 lessons of the module

They come after the prompting method and before the automations. You choose one tool, and the path keeps only that one.

  1. 01 Where ChatGPT lives, and what each door can do 6 min
  2. 02 Generating an image in the conversation 5 min
  3. 03 The custom assistant: building one, and what it won't do 6 min
  4. 04 Permanent instructions and memory 6 min
  5. 05 Reading an image, analysing a file, writing side by side 6 min
  6. 06 Searching the web 5 min
  7. 07 Day-to-day tips 6 min

Honestly

What ChatGPT does better, and where it struggles

A course that only praises its tool is worthless on the day the tool lets you down.

Where it leads

  • The most widespread: your contacts have it, your suppliers have it, the examples circulate.
  • Custom GPTs, shareable across a team without writing a line of code.
  • Image generation inside the conversation, which Claude does not do.
  • Web search, fast and sourced, when you remember to switch it on.

Where it struggles

  • Memory surprises people: it remembers what you thought was gone, and forgets what you thought was set.
  • The free tier opens almost none of what this page is about.
  • Default settings produce chatty answers, and almost nobody changes them.
  • It reaches less far into your own data than Claude, or than Copilot inside Microsoft 365.

It may not be your tool

The path is chosen at the fourth module, and it can be changed. If you already spend the day inside another suite, start there.

Claude Projects, connectors and what runs without you, from Anthropic. The Claude course -> Microsoft Copilot Inside Word, Excel, Outlook and Teams, on your own company files. The Microsoft Copilot course -> Gemini Inside Docs, Sheets and Gmail, plus deep research. The Gemini course ->

The full programme

ChatGPT is one stage, not the whole course

What you buy is a course on using artificial intelligence at work. ChatGPT is its most common way in.

Getting started

Understand what you are handling, learn the method, and open the tool you already have.

  • Welcome and diagnostic
  • Understanding generative AI
  • The prompting method
  • Copilot, Gemini, Claude, ChatGPT (one of four) you are here
Getting good

Go further in the request, judge what comes out, protect your data, and do your real work with it.

  • Advanced prompting
  • Judging an answer
  • Your data and the rules of use
  • Writing, translating, summarising
  • Beyond text: image, video, audio
  • Business use cases
  • Figures and spreadsheets
  • Meetings, minutes and messages
  • Serious research and monitoring
  • Presenting and convincing
Putting it to work

Plug AI into your tools with an MCP server, build chains that run without you, and frame the team that uses them.

  • Claude, for automations
  • Agents and automations
  • Building your MCP base
  • Advanced use cases
  • The foundation: making it last
  • Managing a team that uses AI
  • Putting it to work

Pricing

One course, two months of access

The full path, your tool included, the assistant that answers while you learn, and the certificate at the end. Nothing is billed monthly.

You can add the automations later, without paying twice for what you already have.

The method

290 €

  • 82 lessons, 8h 40min
  • The ChatGPT path, 7 lessons
  • 18 exercises reviewed
  • 60 days of access
Start now

amounts read from the application

Questions

What people ask us about the ChatGPT course

Is the free version of ChatGPT enough to follow the course? +

For the first lessons, yes. Custom GPTs, deep research and file analysis need a paid plan with OpenAI, which we do not sell and which does not go through us. The course says when that becomes necessary, and what you can do without it.

I have used ChatGPT for a year. What is left to learn? +

That is the most common profile in our sessions, and the one that gains the most. What changes is not the tool, it is how you ask: the five parts of a request, recovering an answer that missed, judging what comes out. Permanent settings, custom GPTs and sourced search do the rest.

How much time does it take? +

Around 8h 40min of reading and exercises across the path, spread however you like over your access period. Most learners do one lesson a morning, over about thirty days.

Is this online or in a classroom? +

Online, at your own pace, with exercises reviewed by a human. If you want a live session, in a room or remote, for a whole team, that is a different offer and we run those too.

Is there a certificate? +

Yes, a named certificate at the end of the path, listing the areas covered and the volume actually completed. It is not a diploma and we do not pretend otherwise.