Workshops
Where are you on your journey?
Starting out? Levelling up? Or looking for something advanced?
Below is a selection of workshops I deliver regularly. Each one is tailored before delivery to your organisation’s approved platforms, policies, roles and real workplace examples. Need something more specific? I also design bespoke sessions from scratch.
Tap a workshop to see who it’s for, what it covers and what your team will learn.
-
Audience: General staff, mixed AI experience.
Entry-level, flagship session covering how LLMs actually work, what they get wrong, the AI landscape, data security and safe use, and prompting fundamentals.
What participants will learn
Explain why AI can produce convincing but incorrect answers.
Apply a practical structure to improve prompts.
Identify information that should not be entered into unapproved tools.
Use AI more effectively for common workplace tasks.
Recognise when human review and verification are required.
-
Audience: General staff, mixed AI experience, with no assumed exposure to agents or automation.
A short, practical session on recognising when AI use has tipped into an actual security or data problem, and what to do in the moment. A companion to Foundations: Foundations covers safe use going in; this session covers what to notice and do when something has gone wrong.
What participants will learn
Recognise the kinds of mistakes that turn into a real security or data incident, not just a poor-quality output.
Identify AI-generated content designed to look trustworthy, including phishing, fabricated details and impersonation.
Know what to do in the first few minutes after spotting a problem.
Understand why this is not only IT’s responsibility anymore.
Know the organisation’s actual reporting process, tailored per client.
Tailoring: This session is tailored to your organisation’s escalation process. A short briefing beforehand covers your incident reporting chain, relevant policies and who staff should contact. Without this, the reporting section defaults to generic best practice.
-
Audience: Staff who understand the basics and want to use one approved platform more effectively.
A platform-specific deep dive using realistic workplace tasks rather than a general product tour. Delivered as a separate, focused session per platform, tailored to your organisation’s account type and data controls.
What participants will learn
Navigate the platform confidently and understand its main workplace features.
Choose the most appropriate feature or mode for a given task.
Organise files, instructions and conversations for repeatable tasks.
Understand relevant privacy, account and data-control settings.
Create a simple reusable assistant, project or workspace where supported.
Recognise the platform’s limitations and when another approach is more suitable.
-
Audience: General staff, mixed AI experience.
A hands-on follow-on to Foundations. Less theory, much more doing. Participants draft real emails, structure meeting notes, brainstorm and build reusable workflows using their own work examples.
What participants will learn
Use AI to draft and improve workplace emails and messages.
Turn rough notes or transcripts into structured summaries, actions and follow-ups.
Use AI to brainstorm options and pressure-test ideas.
Improve weak first drafts through follow-up prompts and iteration.
Create reusable prompts, instructions or project spaces for recurring tasks.
-
Audience: Staff who regularly use spreadsheets for reporting, reconciliation, administration or data preparation.
A framework for spotting spreadsheet automation opportunities, then using AI to help write formulas, scripts and repeatable workflows. Delivered in an Excel-led or Google Sheets-led version, covering how to choose between built-in features, formulas, data-transformation tools and scripts.
What participants will learn
Identify spreadsheet tasks that are suitable for automation.
Describe a problem clearly enough to identify and build an appropriate solution.
Provide AI with the relevant structure, rules and edge cases.
Test AI-generated scripts safely using a copy of the original file.
Troubleshoot errors by giving AI useful feedback.
Recognise when a task requires IT support or human judgement.
Tailoring: The Excel or Google Sheets version can be adapted in advance to match your organisation’s environment.
-
Audience: Staff involved in research, policy, compliance, reporting, communications or decision support.
A framework for summarising accurately, running research safely and synthesising across multiple sources, with a strong emphasis on verification given the hallucination risk in this kind of work.
What participants will learn
Select an appropriate approach and tool for source-grounded analysis, multi-document synthesis or wider research.
Write prompts that produce focused summaries for a defined audience.
Compare and synthesise information across several sources.
Distinguish between source-grounded analysis and general AI-generated research.
Trace important claims back to their original source.
Identify weak citations, unsupported claims and likely hallucinations.
-
Audience: Staff who already understand prompting fundamentals and regularly use generative AI at work.
A follow-on workshop for confident users who want more consistent, structured and repeatable results by breaking complex tasks into stages, controlling output quality and building reusable workflows.
What participants will learn
Decompose complex tasks into clear stages.
Create structured and reusable prompt templates.
Use examples, constraints and evaluation criteria effectively.
Separate generation from critique and revision.
Test workflows across several scenarios.
Maintain appropriate human review and judgement throughout.
-
Audience: Regular AI users comfortable with prompting fundamentals.
Moves beyond wording better prompts to giving AI the right information, instructions and reference material from the start. Participants learn how to set up projects, custom instructions and knowledge files so AI produces consistent, relevant results without starting from scratch each time.
What participants will learn
Explain why the context AI is given matters as much as how the prompt is worded.
Identify the background, examples and constraints a task actually needs.
Set up reusable projects, custom instructions or workspaces with the right reference material.
Structure documents and knowledge files so AI can use them reliably.
Manage long conversations and recognise when to reset or refresh context.
Keep sensitive information out of shared or persistent context.
-
Audience: Staff who have completed AI Foundations, or have equivalent experience.
A practical introduction to AI agents: what separates an agent from a chatbot, how to spot suitable use cases, how to write a clear agent brief, and how to build and test a low-risk prototype on one selected platform.
What participants will learn
Explain the difference between a chatbot, reusable assistant, automated workflow and AI agent.
Identify tasks suitable for agent-based automation, and recognise tasks that are too risky.
Write a clear agent brief covering goals, tools, boundaries and escalation.
Build and test a simple prototype using the selected platform.
Apply appropriate guardrails and human review before allowing an agent to take action.
-
Audience: Experienced AI users who have completed AI Agents: From Idea to Working Prototype, or have equivalent experience.
A half-day, hands-on session on designing more advanced agent workflows. Covers when a single agent is not enough, how to chain steps together, build review and retry loops, and coordinate multiple specialised agents, with a strong focus on reliability, oversight and knowing when simpler is better.
What participants will learn
Recognise when a task needs a multi-step or multi-agent approach, and when a simpler setup will do.
Design chained workflows that pass outputs cleanly from one step to the next.
Build review, correction and retry loops with clear stopping conditions.
Map roles and handoffs between specialised agents.
Add checkpoints, logging and human approval at the right points.
Test, troubleshoot and assess the cost and risk of a multi-agent workflow.
-
Audience: Managers and team leaders introducing AI and supporting their teams through adoption.
A practical session on leading AI adoption within a team: setting realistic expectations, choosing good starting points, addressing concerns openly and building the habits that help AI use stick, safely and consistently.
What participants will learn
Build a realistic picture of what AI can and cannot do for their team.
Identify high-value, low-risk starting points for team adoption.
Respond to common staff concerns about job impact, quality and trust.
Set clear team expectations for approved tools, data handling and human review.
Support team members with different levels of confidence and skill.
Track adoption and judge whether AI is delivering real value.
-
Audience: Leaders and managers who approve AI-assisted work, manage risk or set responsible-use expectations.
A plain-language introduction to AI governance for people who approve and oversee AI use rather than write policy. Covers the key risks, the regulatory landscape including the EU AI Act, and the practical questions to ask before approving AI use in a team or process.
What participants will learn
Understand the main categories of AI risk: accuracy, bias, privacy, security and transparency.
Understand the practical implications of the EU AI Act and GDPR at a non-specialist level.
Apply a simple set of questions when reviewing a proposed AI use case.
Set proportionate expectations for human oversight and accountability.
Recognise when to involve legal, data protection or IT specialists.
Contribute to or review a team-level responsible-use policy.
Still unsure?
Thomas ran a series of AI workshops with my teams that were practical from the start. They were built around the work we actually do, and opened the teams’ eyes to how much more the tools could help them with. Everyone came away more confident about bringing AI into their workflows safely.
Brian McCarthy, Group Commercial and Operations Director, Onic
Not sure which workshop fits?
Tell me a bit about your team and I’ll email you a recommendation. No call needed.
What happens next
I’ll read your answers and reply within one working day with the workshop I’d suggest, and why. If none of them fit, I’ll tell you what a bespoke session could look like.
No sales sequence, no follow-up calls unless you ask for one.
Prefer to talk it through?