IPPTA Business Meeting 2026 · International Private Physiotherapy Association · 9 September 2026
Full-day workshop
Overview
A full-day, hands-on workshop for private practice owners at the International Private Physiotherapy Association’s business meeting in Hamburg, run as four 90-minute sessions that each build on the one before it.
The premise is that practice owners already have the expertise; what they don’t have is the time to turn it into content, communication, and a plan. So the day is built around production rather than explanation. Delegates work on their own practice throughout, keeping everything they produce in a single plain-text practice file that carries context between sessions — the answer to the context window, which is the one technical idea the day insists on.
The sessions:
- Foundations — how generative AI works, what the context window is and why it governs everything else, what to keep out of it, and the structure of a prompt (role, context, task, constraints, output) arrived at by building one by hand rather than by being taught it
- Creating content — naming real client groups, building one out as a persona with the model filling gaps rather than inventing the customer, drafting a newsletter for that one person, teaching the model your voice, and spinning the piece into five formats
- Getting found and trusted — reviewing a practice website as a prospective client in pain, an AI-use policy clients can read, a staff guide for the conversation that starts “ChatGPT told me…”, and a communication plan with an owner and a measure on every row
- Growing the business — a walkthrough of a completed Deep Research report, then delegates run their own (or take the interview route on tools without it), closing with the most important prompt of the day: now argue against it
Everything demonstrated runs on free tiers. The worked examples come from a real Lincolnshire practice website and a fictional practice plan, and the day closes with an AI-generated handover report so delegates leave with one page they can act on.
After the day
Participants left with a single plain-text file holding everything they built during the day: a description of the practice, a client persona, a newsletter written for that person, an AI-use policy, a staff guide, a communication plan, and a draft three-year plan. That file is the thing to keep. Pasting it into a new conversation gives the model the context it needs to start where the last session ended, and adding to it after each piece of work means it gets more useful over time.
The last exercise of the day was to ask the model to argue against the plan it had just written. I’d suggest doing this with anything the model produces that you’re about to act on. It knows nothing about your town, your staff, or your lease, and the quickest way to find the gaps is to ask it to look for them.
Everything in the workshop ran on free accounts, with one exception: Deep Research needs a paid subscription, and we covered an interview-based alternative for anyone without one. A paid account is worth having when a limit is stopping you from doing work you want to do, and not before.
Handout
The one-page handout covers the structure of the day, three concepts it depended on (the context window, sycophancy, and personas), four things to do next, and where to read more. It’s written for anyone running a practice, so you don’t need to have been in the room to use it.
Further reading
I’ve written about the ideas behind the workshop on this site. Most of these posts are addressed to educators, but the argument in each is the same one we worked through in Hamburg.
- Context engineering for educators explains why what you give the model to work with matters more than how you phrase the request. The practice file is a small example of this.
- Building AI personas for professional practice is the longer version of the teach it your voice exercise: writing down what you believe and how you work, so that the model can work within it.
- The hidden inefficiency in how we work with AI is about choosing a model for the task rather than treating them all as the same tool.
- Organising your notes for AI covers why plain text works, and what gets lost when you organise information for a machine to read.
Slides
Use arrow keys or click to navigate slides. Open slides in full screen