Also available separately / AI training
Understand AI before applying it at work.
Training is a standalone engagement, even with no software project under way. We work on what AI can do, its limits, sensitive data and checking answers, with examples close to the company’s everyday activities. Remotely or on site, anywhere in Italy.
01 / When it helps
When it can help.
There is no shared language
People hear about AI, but have no shared criteria for telling a useful application from a claim that cannot be verified.
Experiments are scattered
Prompts, tools and habits emerge piecemeal, without clear guidance on sensitive data, sources and human review.
You need guidance before starting a project
The company wants to identify which activities are worth examining, without jumping straight to a solution to be developed.
02 / See it working
Practise with a recurring request, without automating it.
Illustrative example · not a client case study
- People remain responsible
- Interpreting the case, the final decision and external communication.
The exercises use fictional requests and approved procedures, with no personal or confidential data. The aim is not to automate, but to learn how to use and check.
Request What arrives
A fictional request and an approved procedure, with no personal or confidential data.
Draft What it proposes
A draft summary and a checklist to discuss together.
Check Who checks
People check sources, omissions and tone before deciding whether the draft is useful.
Case flagged If something is off
If the request is ambiguous or asks for unauthorised information, the exercise stops and the case is flagged.
03 / How I work
Build a shared understanding through practical examples.
I adapt the programme to the roles involved and the work currently done, keeping learning, process analysis and development separate.
- 01
Start with everyday activities
I start from everyday examples and practical questions from the people involved.
- 02
Clarify what AI can and cannot do
I separate what a model can assist with from what calls for judgement, accountability or better data.
- 03
Try AI with guidance
The exercises may use OpenAI, without making it the centre of the training.
- 04
Practise checking
Checking sources, spotting errors, handling sensitive data and handing over to a person are all part of the exercise.
- 05
Identify the activities
I describe the activities to examine, with the reasons and the questions still to be answered.
04 / What can remain
Define content and materials.
Content and materials are defined once the company’s roles, available tools and objectives are understood. There is no single programme that fits every context.
Practical sessions and slides
Explanations and examples built around the participants’ activities, the limits of the models and how to check their answers.
Safety guidelines and starter prompts
Guidance on sensitive data, sources and human review, together with a carefully chosen set of prompts to start from.
Exercises and activities to analyse
Guided exercises and an initial list of processes for separate analysis, with reasons for each choice.
05 / Questions and boundaries
Frequently asked questions.
Is it a certification?
No. It does not attest to competence or compliance.
Does it replace legal or privacy advice?
No. Policies, privacy and security remain with qualified professionals.
Which tools do we use?
Products, access and materials are agreed beforehand; the exercises may use OpenAI without making it the centre of the training.
What comes after the training?
Process analysis maps out an activity you have already identified before choosing a course of action.
What would you like to be able to do with AI?
To define the scope, an outline of the roles involved, tools already in use, recurring questions and required precautions is enough.