← Back to workshops
Workshop 02 Advanced Full day

Advanced AI in practice

Workshop 01 gives you the foundations. This session goes deeper into the skills that matter day to day: writing prompts that work reliably, understanding what different LLMs are good at, when to use which tool, how to spot weak output, and how to handle data sensibly across providers. Still plain English — not a developer course — but substantive enough that your team stops guessing.

In person or remote Up to 12 people

Who it's for

  • Team members who use AI regularly and want to get consistently better results
  • Managers deciding which AI tools to standardise on across the business
  • People who will write prompts, review AI output, or train colleagues

Agenda

9:00 AM

Prompt engineering that works in business

Structure, context, constraints, and examples. Why vague prompts produce vague output — and how to write prompts your colleagues can reuse.

10:00 AM

Different LLMs — strengths, trade-offs, and cost

What separates models in plain terms: reasoning, speed, context length, multimodal, cost. When to reach for which — and when a simpler tool is enough.

11:00 AM

Break

11:15 AM

System prompts, templates, and reusable patterns

Build a starter library for your business — tone of voice, output format, review checklists. Patterns for drafting, summarising, extracting, and assessing.

12:15 PM

Lunch

1:15 PM

Evaluating output — spotting hallucinations and weak answers

Red flags, verification habits, and when to regenerate vs rewrite. What "good enough" looks like for different task types.

2:15 PM

Data, privacy, and provider choices

What can go into which tool, enterprise vs consumer accounts, retention settings, and agreeing sensible rules for your team.

3:15 PM

Apply to your tasks — workshop and next steps

Rewrite prompts for real tasks from the room. Agree a shared prompt library and model guide for the next 90 days.

You leave with

  • A reusable prompt library with templates for your most common tasks
  • A plain-English guide to which LLMs and tools fit which jobs in your business
  • Shared rules for data handling and output review
  • Confidence evaluating AI output — not accepting the first answer by default