Communication
+ AI.
Speaking Skills + AI (Claude)
presentation,
and AI skills.
Your AI is only as good as your communication. Most teams plateau within weeks — not because of the tool, but because they were never taught to direct it. This crashcourse closes that gap.
Most teams plateau with AI after three weeks.
The first weeks are exciting. ChatGPT writes emails. Claude drafts a brief. The team feels productive. Then output quality stops improving. Some people quietly stop using it. Others churn through prompts hoping something sticks.
The plateau isn't a tooling problem. It's a communication problem. AI rewards clear thinkers who structure intent precisely. It punishes vague briefs by producing vague output at scale. Most teams haven't been trained to communicate clearly — and that gap shows up the moment they start working with AI.
This crashcourse closes the gap. The same principles that turn a working presenter into a real one — specificity, intent, contrast, structure — turn out to be the same principles AI rewards. Train one, you've trained both.
Dylan Smith — nine years of formal actor training (Royal Central School, London), three years as expert Claude operator since launch. Based in an Amsterdam AI startup village, working daily with the tools you're trying to master.
Six things you'll actually use on Monday.
Communication first
Why the principles of clear presenting (intent, structure, audience, contrast) are exactly the principles AI rewards. Train one, train both.
Prompt design
The structure that turns vague requests into reliable, repeatable outputs. Drop-in templates you can use across Claude, ChatGPT, or any tool.
Agent building
Watch one custom agent built end-to-end — live. The decisions, the failure modes, and how to debug a Claude project from scratch.
Workflow optimization
How to identify the workflows worth automating, the ones to leave alone, and the realistic time-savings to expect from AI integration.
Common failure modes
The five most common ways teams fail with AI — and how each one traces back to a communication breakdown that has a fix.
Your prompts, debugged
Bring real prompts, real tasks, real frustrations from your own work. Leave with working solutions and a method you can repeat.
Online or in the room.
The online version delivers the full method in 90 minutes — joinable from anywhere. The IRL Amsterdam version doubles the time and adds the hands-on prompt clinic with the room.
The full method, compressed.
- The SPEAK Method walkthrough
- Live agent build demonstration
- Prompt design templates you take home
- Q&A on your specific use cases
- First Wednesday of every month · 15:00–16:30
- Bring your laptop, work along
The full method + the room.
- Everything in the online version, doubled in depth
- Hands-on prompt clinic — bring real use cases
- Live agent build with real Q&A from the room
- 30 minutes of structured debug session
- Date TBA · Science Park Amsterdam
- Waitlist on the For Individuals page
Three blocks. One outcome.
-
Block 1
Why your AI is plateauing
Diagnostic of where teams get stuck and why. Live demonstration of the same prompt with three different communication levels — and the difference in output quality.
-
Block 2
Prompt design + Live agent build
The full prompt-design framework. Then we build one custom agent live, end-to-end, with the room watching every decision and learning from the mistakes.
-
Block 3
Workflow optimization + Q&A
Bring your real workflows. We work through them with the room — applying the method to each one. By the end you'll have working prompts and clear next steps you can use Monday.
Ready when you are.
For individuals, both formats live on the For Individuals page — the online webinar is bookable directly, the IRL version is via waitlist. For corporate teams, this same content scales to a tailored in-house keynote.
For individuals — book online or join IRL waitlist