Training

From Novice to Expert in One Day

A full-day intensive that takes you from basic AI familiarity to genuine proficiency. Demonstration-led learning with real-world application.

£900/day + £100/attendee

Book Training
The February 2026 Moment

Getting Ahead of the Curve

The shift from "AI chatbots" to "AI that does work" has happened. Cowork, Codex, and agentic tools have fundamentally changed what's possible. Most people are still catching up—this training puts you ahead.

Our approach is demonstration-led: you'll observe complete workflows, see real tasks transformed, and understand the principles through watching them in action. This isn't death by PowerPoint—it's watching an expert work and understanding why it works.

By the end of the day, you'll understand how modern AI actually functions, know how to provide effective context, and be able to delegate complex tasks to AI as a capable worker—not just ask it questions like a search engine.

Full day (7–8 hours including breaks)
Remote or in-person
Demonstration-led with discussion

Course Programme

Seven sessions covering the complete AI workflow

Morning Session

09:00–09:30

Opening: The February 2026 Moment

Why this training, why now

We start with a demonstration—not a slide deck. A chaotic folder of files becomes an organised summary report in minutes. This is what's now possible. Then we frame what you'll be able to do by end of day.

What's changed Agentic AI Setting expectations
09:30–10:30

Session 1: How LLMs Actually Work

Demystifying the technology without getting technical

LLMs as prediction engines, why they're good at what they're good at, and why they make the mistakes they make. Understanding the mechanism helps you work with the tool effectively and predict when it will struggle.

"Understanding why AI hallucinates helps you know when to trust it."
Next-token prediction Training process Hallucinations Capabilities & limits
10:30–10:45

Break

10:45–11:45

Session 2: Context—The Most Important Concept

Everything is context

The context window, what goes into it, and why it matters. System prompts, conversation history, files, memory. Why long conversations degrade and when to start fresh. Markdown as the universal language of AI.

"The quality of AI output is bounded by the quality of context you provide."
Context window System prompts Context management Markdown
11:45–12:45

Session 3: The Art of Prompting

From asking questions to specifying outcomes

The evolution from prompt engineering to context engineering to delegation. The framework: Role + Task + Context + Format + Constraints. Meta-prompting: getting AI to write and improve its own prompts.

"Don't describe the process. Describe the outcome."
Prompt structure Meta-prompting The three questions Iterative refinement
12:45–13:45

Lunch Break

Afternoon Session

13:45–14:45

Session 4: Files—The Unit of Work

From chat messages to deliverables

The paradigm shift from chat to files. Why files matter: persistence, shareability, actual work product. Input files for grounding, output files as polished deliverables. Excel with working formulas, PowerPoint creation, document generation.

"If you're copy-pasting from chat, you're doing it wrong."
File input/output Skills Grounding Deliverables
14:45–15:45

Session 5: Delegation and the AI-First Philosophy

Treating AI as a capable worker, not a search engine

The uncomfortable truth: AI is already better than you at most knowledge tasks. Delegation as a skill. Preparation and research matter more now because execution is cheap. Quality verification when you're not an expert. When NOT to use AI.

"Cowork is a worker, not a chatbot. Delegate outcomes, step away, come back to finished files."
AI-first mindset Delegation skills Quality verification Human judgment
15:45–16:00

Break

16:00–17:00

Session 6: Advanced Patterns

How the most effective users work

Skills: reusable procedures AI discovers and applies automatically. Subagents: when AI spawns its own helpers. Claude.md and AGENTS.md for persistent memory. Code generation for non-coders. MCP and connectors for tool integration.

"Build once, reuse everywhere. Your procedures become AI capabilities."
Skills Subagents Claude.md MCP connectors
17:00–17:30

Session 7: Putting It Together

From theory to daily practice

Building personal workflows, the preparation → delegation → verification cycle, and continuous learning as tools evolve. Interactive discussion: what tasks from your actual work could be transformed?

Personal workflows Real applications Q&A
17:30–17:45

Closing

Recap of key concepts, first steps when you get back to work, and resources for ongoing development.

Key Themes Woven Throughout

1
Outcomes over process Always define what "done" looks like
2
Context is everything Rich input → rich output
3
AI-first, human-verified Let AI draft, you review and refine
4
Preparation is the new execution Thinking is your job, doing is cheap
5
Files, not chat Work in deliverables, not conversations

What You'll Take Away

Deep Understanding

Not just how to use AI, but why it works the way it does. You'll predict when it will struggle and know how to work around limitations.

Delegation Skills

Treat AI as a capable worker. Delegate complete tasks, step away, and come back to finished deliverables—not endless back-and-forth chat.

File-Based Workflows

Move beyond chat to real work product. Upload source material, receive polished documents, spreadsheets with working formulas, presentations.

Confident Judgement

Know when AI is the right tool and when it isn't. Verify quality even when you're not an expert in the domain.

Ready to get started?

Tell us about your team and we'll arrange a session that fits your schedule.

Book Training