- Question
- How to apply AI to the team's own recurring tasks
- Output
- A written method for each task covered
- Who it suits
- Executive teams, departments and functional teams that have AI tools but use them unevenly
- Duration
- From a 60 to 90-minute executive briefing to a 4-week program
- Fee
- We quote the fee for each workshop
- Does not include
- An organisational audit or a strategy roadmap
Who it suits
The Practical AI Workshop is for executive teams, departments and functional teams that have AI tools but use them unevenly. In these teams a few people use the tools every day. Others tried them once and went back to the old way of working.
The workshop shows how to apply AI to the team's own recurring tasks. Each workshop needs a defined work outcome and a manager accountable for it. The outcome might be faster first drafts of client proposals, a shorter month-end commentary cycle or a consistent method for preparing board papers. The manager makes sure the new practice is still in use after the workshop.
A leadership team can also take the workshop itself, to try the methods before it asks its people to. Staff are more likely to use the tools when leaders use them openly. Ramon develops this point in his article on culture and AI adoption.
The workshop fits when:
- a manager will own the outcome and follow up on use
- the tasks recur and people can bring real examples
- the leaders who will ask for the change are willing to take part, or already have.
Another starting point may suit you better when:
- your executives want coaching on their own papers and decisions: see Executive AI Coaching
- you need to decide where to invest in AI: start with the AI Opportunity Audit
- the first decision is which AI tool to approve: start with the AI Roadmap
- one workflow is failing and needs fixing: start with the Operational AI Audit
- you want a talk for a conference or an all-staff event: see the speaking page.
Formats and duration
Four formats are available, from a 60 to 90-minute executive briefing to a 4-week program.
| Format | Duration | Suits | What people leave with |
|---|---|---|---|
| Executive briefing | 60 to 90 minutes | An executive team or board that needs a shared view before it sets direction | A common understanding of what AI can and cannot do in your business, and the decisions it raises for the team |
| Half-day workshop | Half a day | One team that shares recurring tasks | Hands-on practice on the team's own tasks, and a written method for each task covered |
| Full-day workshop | 1 day | A team or function with several kinds of work, or work that needs closer checking | Practice across more tasks, written methods and agreed checking rules for sensitive or consequential work |
| Multi-week program | 4 weeks, 1 session a week | A team that needs the new practice to hold through a full work cycle | Methods applied to live work between sessions, with each session reviewing what was used and what changed |
In the multi-week program, the first session covers the half-day content. The second and third sessions each take one task type onto live work. The fourth reviews use with the manager and finalises the written methods.
We travel to you for half-day and full-day workshops, where people work through tasks together. We agree group size with you when we scope the workshop.
What happens in the room
Your people work on tasks from their own roles, with the tools you have approved and the controls that apply to them. Each exercise starts from a task your people already do. Ramon Rodriguez leads each workshop.
A half-day workshop runs in 5 parts. The other formats follow the same sequence, with the depth set by the time available.
| Part | What happens |
|---|---|
| 1. Choose the work | Each person brings a recurring task from their role, and the group agrees what a good result looks like for each one. |
| 2. Write the brief | People state the decision or action the output must support, who will use it and what would make it unusable. They add what the tool cannot know: the audience, the constraints, the decisions already made and the sources it may use. |
| 3. Work in stages | The task is split into steps: establish the facts, test the assumptions, form the recommendation, then write it up. People review the reasoning before anything is formatted. |
| 4. Check and correct | People check the output against the original sources, name the specific defects and correct them. The group agrees which work needs a named person's approval before it leaves the team. Examples include public statements, legal interpretations, customer commitments and capital decisions. |
| 5. Agree the practice | Each person records the method that worked and when they will next use it. The manager agrees how use will be followed up. |
A model predicts a likely answer from the information it is given. It does not know your business, and it can sound confident when it is wrong. Most of the time in the room goes to the brief and the checks, because they control the quality of the result. Ramon explains the reasoning behind this sequence in Why capable AI still produces average work.
A brief written in the room might read: 'Decide whether this proposal is ready for executive approval. Test the commercial assumptions, the implementation ownership and the evidence of value. Separate established facts from forecasts. Recommend approve, revise or stop.'
Tools covered
The workshop covers Microsoft Copilot (formerly Microsoft 365 Copilot), ChatGPT, Claude and Gemini, and uses the one your organisation has approved. People practise in the tool they will use at work, under the settings your organisation has configured.
The same underlying model can behave differently in different products. Each product decides how much reasoning the model applies and which information it can reach. People learn the relevant settings in their own tool. Where more than one tool is approved, they learn which tasks suit each.
If you have not yet approved a tool, the workshop can run on the one you intend to approve, using material with no confidential content.
Before, during and after
- Before: A conversation with the accountable manager sets the work outcome, who attends, and the tools and controls that apply. We ask for examples of the recurring tasks people will bring, with confidential detail removed.
- During: People work on their own tasks in the approved tool. They enter only the information your policy allows in that tool.
- After: The manager follows up on use. If a task now takes less time, the manager decides what work that time moves to. Our note on defending an AI time saving sets out how to report the time saved. An optional review session a few weeks after the workshop checks which practices are in use and adjusts the rest. We quote it with the workshop.
What your team leaves with
Each person leaves with a written method for each task covered. Each method records the task, the brief that worked, the context to supply, the checks to run and who approves the result.
The team keeps the methods where people can find and reuse them. It also gets advice on where coaching or champions would help next.
Some tasks will not improve with AI yet. The workshop records those as well, so people stop spending time on them.
The accountable manager also receives a one-page written summary of the workshop. It records the tasks covered, where the new methods worked and what would help next.
What it does not include
The workshop does not include an organisational audit or a strategy roadmap. If it raises a workflow or investment question, we scope that as the next stage. Applied AI Australia advises and Acquire Intelligence™ delivers, as one engagement from diagnosis to working change.
What we need from you
We need:
- a named manager accountable for the work outcome
- attendees who can bring real examples of their recurring work
- working access to the approved tool for every attendee on the day
- your organisation's AI policy or usage rules, so the exercises follow them
- time for the manager to follow up on use after the workshop.
How the fee is set
We quote the fee for each workshop. Four things set it: the format, the number of people, whether the sessions run on site and the preparation needed. Preparation covers building the exercises around your roles and tools. We quote a multi-week program for its full length. You receive the fee in writing before you commit.
Frequently asked questions
Which format should we start with?
Start with an executive briefing if your leaders have not yet used AI on their own work. It gives them a shared view before they ask their teams to change. Otherwise, a team that shares recurring tasks suits a half-day or full-day workshop. A multi-week program suits a team that needs the practice to hold through a full work cycle.
Will senior people be hands-on?
Senior people work hands-on, on a task from their own role, in the tool they will use. They take part like everyone else.
Is the workshop built around our own work?
The workshop is built around your own work: people bring recurring tasks from their roles, with confidential detail removed. A model fills any gap in a brief with the most common pattern it has seen. Real tasks show people where their own briefs have gaps.
Which AI tools does it cover?
It covers Microsoft Copilot, ChatGPT, Claude and Gemini. People practise in whichever of them your organisation has approved, under its configured settings.
Can it run online or on site?
Half-day and full-day workshops run on site, while the briefing and the multi-week program's sessions can run online. Preparation runs remotely for every format.
What happens after the workshop?
The accountable manager follows up on use and decides where any time saved goes. Each person keeps the written method for their task. A workflow problem or investment question the workshop exposes becomes the next stage, under Resolve a business problem. Executives who want to go further on their own work can move to Executive AI Coaching.
How is confidential information handled?
People use only the tools and settings your organisation has approved, and bring examples with confidential detail removed. Board deliberations, negotiation positions, legal matters and personal information about staff or customers stay out of consumer AI accounts. In their business tiers, the 4 vendors each limit the use of customer content for model training, on their own terms. Microsoft states that Microsoft Copilot prompts, responses and Graph data are not used to train foundation models. OpenAI states that ChatGPT Enterprise, Business and Edu data is not used for training by default. Anthropic states that prompts, data and results on its enterprise offering are not used to train its models by default. Google states that Workspace content is not used for generative AI model training outside your domain without permission. Confirm the settings, retention and data location in your own tenancy before the workshop.
Tell us which team would take part and the work outcome its manager will own.