AI

There’s knowing AI. And knowing what to do with it.

Learn both, with someone who brings engineering, an MBA, teaching and the daily realities of running a business into the same conversation.

I teach practical AI to individuals, in a live small group or 1:1, and work with organisations on team training and on applying AI to real work.

Explore AI learning AI for your organisation

Not sure where to begin? Find your starting point

What’s behind this briefing?

In my own business

Some of the most useful things I’ve built with AI are things you’ll never see.

Enquiries reach my business on WhatsApp and by email, faster than anyone can keep every record up to date. By the morning, AI-assisted systems have read the new conversations and relevant emails, checked them against our pipeline and prepared a short briefing.

The useful part isn’t a shorter version of everything. It’s noticing how a conversation, a record and the next step relate.

Illustrative reconstruction of an internal workflow. Fictional data.

What happened

WhatsAppBusiness number

  • Marta

    Hi! Do you have evening lessons? I finish work at 6…

    Unread 1: used in briefing item 1

  • Chiara D.

    Is a Saturday morning possible for 1:1?

    2: used in briefing item 2

  • Luca B.

    You: Perfect, see you on Thursday!

EmailInbox

Paolo Conti

Re: Course for our team

Hi, a quick update: we’ve now confirmed four people. Could we start in November rather than October? Best, Paolo

3: used in briefing item 3

What the record says

PipelineEnquiries
NameLast contactSuggested update for review
MartaEvening lessons Wed · WhatsAppAdded today New enquiry. Added from WhatsApp. 1: used in briefing item 1
Chiara D.1:1 lessons 26 Sep · WhatsApp No reply recorded. Follow-up drafted, not sent. 2: used in briefing item 2
Paolo C.Team course 30 Sep · Proposal sent Email Wed: four people, November start? 3: used in briefing item 3
Elena V.Group course 2 Oct · Enrolled No action needed.

What needs a person

Morning briefingThu · 07:30

Start here today

Written in pen beside the heading: “Start with 2.”

Checked: WhatsApp, email and the pipeline.

  1. New enquiry

    A prospective learner asked about evening lessons on WhatsApp. Added to the pipeline. A first reply is ready to review.

    From WhatsApp + pipeline

  2. Follow-up to check

    An earlier enquiry has no reply recorded. Check the conversation before using the prepared follow-up.

    From pipeline + WhatsApp

  3. Existing prospect

    New information arrived by email. Review it alongside the existing pipeline record before responding.

    From email + pipeline

Supporting detail · item 1

WhatsApp · Marta · Wednesday

Hi! Do you have evening lessons?

I finish work at 6, so weekdays after that would be ideal.

Pipeline: no existing record matched this number, so a new row was added for Marta (evening lessons, from WhatsApp).

Draft reply · WhatsApp · not sent

Hi Marta, thanks for your message! Yes, we have evening options. As you finish at 6, I’ll check which weekday groups start after 18:30 and send you the choices today. What would you most like to work on?

Built from the conversation: a evening lessons b finishes at 6 c weekdays

A person reads it first, then edits, sends or discards it.

Held back: the follow-up for item 2 waits until someone has checked that conversation.

Numbers show which pieces each item in the briefing was built from. Printed: what the system prepared. In pen: my call.

The system brings the pieces together. I decide where to put my attention.

That is the division of labour. These systems help me understand what is happening in the business. The drafts wait for a person. A gap between a conversation and a record is a reason to look, not a conclusion. Similar summaries cover other parts of the operation.

They also make room for the work I want to do next: websites, educational projects and the next stage of my brand. This website is one of them.

You won’t need to build anything like this to learn with me. The thinking behind it is what transfers: where does your own work arrive, and what would you want to know first each morning?

From a 1:1 session

The problem wasn’t only the prompt.

A professional came to a session disappointed with Copilot at work: the answers weren’t much use. We worked through it together on a shared screen. The wording turned out to be only one part of the problem.

A real session, described anonymously. The wording below is reconstructed for illustration; it is not the original request or output.

01The technology

Was the tool working with the information the task required?

At first

What it could see

  • Relevant email: not available
  • OneDrive material: not available

Not the information this task needed.

After we worked through it

What it could see

  • Relevant email
  • OneDrive material

We found an appropriate working context and set up access to what the task required.

02The business question

What did the team need to clarify about responsibilities?

At first

Create a responsibility matrix for my team.

After we worked through it

Using the recent project emails and the team’s files, show who currently handles each recurring task, where responsibilities overlap or are missing, and which questions the team needs to settle.

A matrix was the format. Unclear responsibility was the problem.

03The useful output

What would help them hold a better-informed discussion?

At first

Plausible, general answers. Nothing specific to this team.

After we worked through it

A working draft to take into the team discussion: who does what now, where it overlaps, where nobody owns it, and what needs deciding.

Useful because it could be tested with the team, not because it looked impressive.

In this case, as I saw it, the meeting became more useful, and they were able to contribute with more influence.

The useful question isn’t only “What should I type?” It’s also “What is this tool seeing, and what am I actually trying to resolve?”

What we learn

Understanding is part of the practical skill.

Alongside the tasks, we look at the wider picture: what AI is, how it reached its current form, how to read the landscape of tools and models, why differences between systems and versions matter, and where the limits are.

Not as a lecture. Each part earns its place by answering a question you will meet in real use.

  1. 01

    When an answer is poor, what should you investigate?

    Rarely just the wording. Look at the task you set, the context the tool can see, the tool itself, and how you are judging the result.

    TaskContextToolEvaluation

  2. 02

    Why might a different tool or model change the approach?

    Tools, models and versions differ in what they can see, do and get right. Understanding the landscape lasts longer than memorising one interface.

    For exampleOne assistant can read your files, another can search the web, another can write and run code. The same request needs a different approach in each.

  3. 03

    Which developments are worth paying attention to?

    We discuss where the technology appears to be heading and what it could change in your work: possibilities to watch, not forecasts presented as facts.

    For exampleAssistants that take actions in your other tools, not only answer questions, and what that changes about checking their work.

How deep we go depends on where you start and the programme you choose. Theory and practice support each other.

Background

Why I approach AI this way.

I trained as a computer-science engineer, took an MBA at IIM Ahmedabad and have spent years helping people understand difficult subjects. I also founded and run an educational business, which gives me real situations to apply and test ideas.

Each part brings its own question to the same conversation. You have already seen all three on this page.

  1. Engineering

    “What can this tool see, and what can it actually do?”

    Understanding the technology and how things connect.

    Seen in the 1:1 session

  2. Business

    “Is this a problem worth solving, and what changes if we do?”

    An MBA and running a business: judging whether an application earns its place.

    Seen in the morning briefing

  3. Teaching

    “Will you understand it well enough to use it without me?”

    Making the understanding accessible, so it becomes your capability rather than my demonstration.

    Seen in the three questions

“…ability to read both business problems and the wider economic and technological context.”

Duccio Neri, Executive Chairman & Co-founder, Philogen Source: LEAD · Executive coaching · excerpt

Where to go next

Two ways to start.

For yourself

Learn AI for yourself

We start from what you understand now and what you want to be able to do.

People often start with

  • A tool they’ve tried that hasn’t been much use yet
  • A task they repeat every week
  • A question about how it all works
  • Live small groupLearn alongside others, with room for your own questions.
  • Personalised 1:1Your pace, your tasks, your tools.

Foundations are welcome. The systems on this page show what’s possible. They aren’t a prerequisite, or what a first course promises to build.

Explore AI learning

For your organisation

AI for your organisation

Training for teams, practical adoption, and opportunities in the processes your people already run.

  1. Team learningA shared, practical way of working with AI, built around your team’s real tasks.
  2. Adoption in practiceWhere AI fits in daily work, and what stays with a person.
  3. Process work scoped separatelyWhen a process needs redesign, we look at how the work is actually done before changing anything.

For leaders: what to adopt, what to question, and how to bring a team with you.

AI for your organisation

Start with a conversation

What would you like to understand, improve or build?

Tell me where you are with AI and what you would like to be able to do. I’ll suggest a sensible starting point: learning for you, training for your team, or a closer look at a process.

Let’s talk

Start with something real.

Or ask where to begin