Three Things You'll Get from This
- 1. An intro to prompt engineering"Write me a blog post" gets you generic garbage. "Write me a blog post on X for an audience of Y, with this tone, and start with a hook that does Z" gets you something usable. That's the difference. I'll show you why prompts fail, how to structure them so they don't, and the four-word hack that changes everything: ask the model to write the prompt for you. Garbage in, garbage out — and most people are putting garbage in without realising it.
- 2. The tools, the use cases, and the risksChatGPT isn't one thing. It's a product with modes — study & learn, agent mode, deep research, image generation, voice, projects, Codex. Most people don't know half of these exist. Then there's Claude, Perplexity, Abacus, and others — each better at different things. I'll walk you through what's out there, what each tool is actually good for, and where the real risks are: hallucinations, confidentiality, over-reliance. You'll leave knowing which tool to reach for on Monday morning — and which ones to be careful with.
- 3. The one thing from your form, built liveYour intake form asks what you'd build if you could only build one thing: the report that takes too long, the workflow you keep meaning to automate, the project you've been putting off. The second hour is planned around that answer. We build it together in ChatGPT, with projects, structured prompts and a custom GPT if the job needs one, and test it on your real work. Not a hypothetical. Yours, working, before we finish.
It's quite amazing what is possible today. I don't think I appreciated how far along the technology is already beyond basic ChatGPT.
What We'll Actually Cover
After you book, you fill in a short intake form: what you do, what you've tried, your top three workflow frustrations, and the one thing you'd build. That's the discovery. I read it, research your role, and plan the second hour around it before we meet.
First 30 minutes: The foundations
In time, AI will meet you where you are. Right now the edge is five or six things, and we do them live inside ChatGPT rather than on slides. How to prompt: role, task, output, and the four-word hack of asking the model to write the prompt. How to manage context: why long threads go bad, when to start a new one, and how projects and custom instructions mean it already knows who you are. The modes most people never find: deep research, agent mode, study & learn, voice, Codex. And which model you're actually talking to.
Next 60 minutes: The art of the possible
Three or four demos planned from your intake form, on your real work. I drive first and show you the tool. I explain what made it work. Then you take over with your own content while I coach from the side. A project for persistent context. A structured prompt you can rerun every week. A custom GPT if the job needs one. We iterate until the output is right. Where another tool does it better (Claude for long writing, Perplexity for research), I'll say so and show you.
Before we finish: Make it stick
Where the files live. How to set this up on your work computer. How to check the output before you send it. What to do when you hit a wall.
You'll leave with three things to try this week, the thing we built, the prompts and projects behind it, the recording, the searchable transcript, and post-session notes with links to everything we used. You'll know what to do on Tuesday. Want the full breakdown? Read what happens in a session.
Who This Is For
- Strategy and operations people who need to turn information into decisions fast. ChatGPT can be your research assistant. Most people don't use it that way.
- Managers and team leads who need to train their teams or create frameworks. You'll learn enough to teach others. You'll know what's possible.
- Content creators — writers, copywriters, marketers. You already know ChatGPT exists. You're using it casually. Proper prompting cuts your time in half and doubles the quality.
- Anyone who gets handed a task and thinks "ChatGPT could help with this" but has no idea how to ask. That's most people.
- People tired of generic group courses that teach you the interface and nothing else. You need someone who actually uses these tools daily and can show you the patterns that work.
How This Works: 90 Minutes
90-minute 1-2-1 co-work session · £699
You book a 90-minute session and fill in a short intake form. We run it on Google Meet or in person in London, your choice. Thirty minutes on the foundations, sixty on your own work. That's it.
No packages. No upsell. No "you'll need the premium bundle."
- The one thing from your form, built on your real work
- The prompts, projects and files we build, yours to keep
- A recording and transcript of the whole session
- Post-session notes you can refer back to whenever you need them
Available on Google Meet worldwide or in-person in London.
Or if you want to know what happens in a session before you commit, read the detailed walkthrough at what happens in a session.
Questions first? Email me or check out my broader personal AI training page. You can also read about why 1:1 training beats generic AI courses. If your company has rolled out Microsoft Copilot instead, see the Copilot training session.
Common Questions
Do I need a ChatGPT Plus subscription before the session?
No. I have ChatGPT Plus and all the other subscriptions — we'll use mine during the session. You might want to buy it afterwards, and I'll tell you honestly whether it's worth it for your specific situation. Some people get by with the free tier and other tools. Some need Plus. We'll figure that out together.
Will we cover other AI tools, or just ChatGPT?
We'll cover ChatGPT (including the newer modes most people don't know about — agent mode, deep research, study & learn, Codex), Claude, Perplexity, and more — whatever's relevant to your work. ChatGPT is the starting point, not the finish line. I'll show you why some tasks are better on Claude. Why Perplexity is brilliant for research. And why Abacus — the Kayak of AI, one subscription for every model — might save you three separate payments.
What if I'm a complete beginner with AI?
This works for you too. I'll start with the fundamentals. We're not doing computer science. We're talking about how to use these tools. If you've never opened ChatGPT, you can still get value here — but ideally you've played with it a bit. Even 10 minutes.
Do I need to prepare anything?
Just the intake form, about ten minutes after you book: your role, the tools you've tried, your top three workflow frustrations, and the one thing you'd build if you could only build one. That's the discovery. I plan the second hour around it. If you already know the task you're stuck on, put it on the form, screenshots and all.
How much does it cost?
A 90-minute 1-2-1 co-work session is £699. That includes the recording, the transcript and the prompts and projects we build. Team sessions start from £2,499.
How long will what I learn stay relevant?
These tools change fast. New models drop. Interfaces shift. The fundamentals don't. How to write a good prompt? That sticks. How to think about model differences? That sticks. I'll send you updates on major changes, and you can book a follow-up session if you want.
Do I need to be in London to do this?
No. We can run it on Google Meet. The session is the same whether you're in London or anywhere else. I'm based in London, but the training works just as well online.

