How long does the assessment take?

It depends on how many people we need to involve and how accessible the data is. Before we start, we agree on the timeframe and clarify what we need from you and your team.

What do you hand over at the end?

A reasoned recommendation on what to do first, or why not to proceed. If the project is justified, we set out what to build, who is responsible, what we need from your business, how to measure the result and the expected running costs. We also document the baseline data and the options we ruled out, with the reasons.

Does the assessment always lead to a project?

No. We assess a potential project, and the conclusion may be that it should not go ahead. If you do not need a new system or the work is not worthwhile, we explain why in writing.

Do you work with businesses that already have a supplier?

Yes, provided we are responsible for implementing and maintaining part of the project. We define our scope and responsibilities at the outset. If the work belongs with your existing supplier, we say so. We do not duplicate what you have or sell you an assessment to hand over to them.

Do you provide AI training?

Not as a standalone service. We show the people involved how the system we build works, where it can make mistakes and how to spot them. This is part of the project, not a separate course.

What do you keep if you decide not to go ahead?

You keep the baseline data and the reasons for the decision. They give you a record of what was considered and why the project stopped, which you can refer to if you revisit the idea.

Where is it usually best to start?

A task that comes up often, follows reasonably stable rules and uses data you already have. It should let you compare the result with the way you work now. Your most important business process is not necessarily the best place to start.

AI consulting

AI consulting with the team that builds and maintains your system

We start with one task in your business and assess whether AI could help. The assessment is part of the project: it sets out what to build, how to measure the result and what running costs to expect. If the project is worthwhile and you decide to proceed, we build and maintain the system. If it is not worthwhile, we explain why in writing.

What our AI consulting covers

Consulting is the first part of a project: we identify the task to address, the data it needs, what will change for the people doing the work and how to measure the result. The assessment also helps you decide whether to go ahead. If you proceed, the same team handles development and maintenance.

The problems to solve

“We need AI across all our processes”

That leaves the first decision open: which task should you start with? Before development begins, we need to narrow the scope and decide how to assess whether the work is worthwhile.

The proposals are hard to assess

A convincing demo does not tell you whether a project will work for your business. You need to know what data it requires, how it will fit into daily work and who will maintain it.

The assessment stops at a list of tools

You pay for an assessment and receive a list of tools and opportunities. You still need to decide who will build the system, what data they need and who will maintain it. Getting the project started is left to you.

What we do

We choose one task to start with

We look at recurring tasks with you, identify where work gets held up and establish what it costs today. We choose one task to assess first and keep a record of the others for later consideration.

We establish the baseline

Together, we record how long the task takes, how often it happens, the errors and the manual steps. Those figures let us compare the work before and after implementation and assess whether the project was worthwhile.

We also say when it is not worth it

Low volumes, unreliable data or a process that keeps changing can make automation cost more than it is worth. If that is the case, we explain why and put it in writing.

The same team handles development and maintenance

The assessment sets out what is worth building, who is responsible, what we need from your business and how to measure the result. If you decide to proceed, we develop and maintain the system.

What we do not do

We do not offer standalone consulting

We do not sell AI strategy as a standalone service or write reports for you to take to another supplier. The assessment helps you decide whether to proceed with a project that we will build and maintain.

We do not deliver lists of tools

A list of platforms leaves you to work out what to do with them. We define the problem first, then choose the tools and explain why they fit the task.

We do not promise savings before measuring

Savings reported by other businesses for different tasks do not show what you could save. We use your own baseline to assess the potential benefit, even when it falls short of your expectations.

The work you and your team still need to do

A useful assessment takes time from the people who know the work. After launch, your business still needs to manage the data, handle exceptions and make the decisions the system cannot. Our maintenance work does not replace that responsibility.

Time from the people doing the work

The people doing the task every day need to make time for the assessment, alongside the decision-makers. They know which steps slow the work down and which exceptions are missing from the documentation.

The baseline data

Someone in your business needs to gather task volumes, completion times and error counts, even if they are approximate. These are the reference points for assessing the result after launch. Without them, you are relying on impressions.

Someone accountable after launch

You need to appoint someone who remains accountable for the result and has time to oversee it. They do not need to be a technical specialist. They do need to manage exceptions, look after data quality and take responsibility for decisions the system cannot make.

Questions to answer before you start

Which tasks recur each week, and how many working hours do they take in total?
What slows down or blocks the work today: missing data, approvals or handovers between teams?
Which errors cost your business money, how often do they happen and how much does each one cost?
What data is already usable, and what would need to be collected from scratch?
Who would be accountable for the result after launch, and how much time could they commit?

Facts and sources

Among Italian companies that considered AI but did not invest, 58.6% cite a lack of adequate skills as a barrier. It is the most frequently cited reason, ahead of a lack of legal clarity (47.3%) and data quality issues (45.2%).

ISTAT, Imprese e ICT — anno 2025, pubblicato il 15 dicembre 2025consulted on 2026-09-20

Questions about AI consulting

How long does the assessment take?: It depends on how many people we need to involve and how accessible the data is. Before we start, we agree on the timeframe and clarify what we need from you and your team.
What do you hand over at the end?: A reasoned recommendation on what to do first, or why not to proceed. If the project is justified, we set out what to build, who is responsible, what we need from your business, how to measure the result and the expected running costs. We also document the baseline data and the options we ruled out, with the reasons.
Does the assessment always lead to a project?: No. We assess a potential project, and the conclusion may be that it should not go ahead. If you do not need a new system or the work is not worthwhile, we explain why in writing.
Do you work with businesses that already have a supplier?: Yes, provided we are responsible for implementing and maintaining part of the project. We define our scope and responsibilities at the outset. If the work belongs with your existing supplier, we say so. We do not duplicate what you have or sell you an assessment to hand over to them.
Do you provide AI training?: Not as a standalone service. We show the people involved how the system we build works, where it can make mistakes and how to spot them. This is part of the project, not a separate course.
What do you keep if you decide not to go ahead?: You keep the baseline data and the reasons for the decision. They give you a record of what was considered and why the project stopped, which you can refer to if you revisit the idea.
Where is it usually best to start?: A task that comes up often, follows reasonably stable rules and uses data you already have. It should let you compare the result with the way you work now. Your most important business process is not necessarily the best place to start.

Tell us which task takes up too much time

Tell us about a task that takes more time each week than it should. We will look at it with you to see whether AI could help or whether a different change to the process would be more useful.