AI Assessment

Who uses AI at your company, for what,

and what it costs you

The answer is usually "more or less". In four weeks we replace it with numbers and a document you can circulate in the boardroom. It helps if AI has already crept in on its own, one personal account at a time, and it helps if it has yet to arrive.

Let's talk

4 weeks · duration and price written in the proposal, before we start

Whether AI is already there or still has to be brought in, a measure is missing

If it has already arrived on its own

  • Everyone uses their own tool, often with a personal account
  • Licences have an invoice, consumption does not
  • Documents and data end up in prompts whose destination nobody knows
  • Some use it well and some use it badly, and nobody knows who

If it has yet to arrive

  • The tool is chosen first and the purpose comes later
  • Use cases are decided by gut feeling
  • Everyone gets the same course
  • There is no starting point, and a year from now nobody will be able to say what has changed

The problem in numbers

The numbers of those who start without a measure

These are public numbers, with the source given below.

73%
Companies overshooting their AI cost plan
2.4x
Average overshoot on agentic projects
$200-600
Real cost per developer per month
2.5-3.5x
Healthy ROI, but only if you count tokens

Public sources: FinOps Foundation State of FinOps 2026, GitClear study on 153 million lines, 2026 productivity benchmarks.

Three sources, set side by side

The questionnaire

To everyone in scope. About 35 closed questions, fifteen minutes, anonymous.

System data

Licences, consoles, invoices, API consumption, where they exist. They show what you pay for and what you actually consume.

The interviews

Half an hour with each function head, on processes, volumes and priorities.

People tell you what they do and what they would like to do, systems show what is consumed. The two often differ, and that is where the work to be done lies.

A score per function, across five dimensions

Each function receives a score from 0 to 100 on adoption, skills, risk, value and costs. Behind every number are the answers that explain it. The table is an example with made-up data.

A score per function, across five dimensions
AdoptionSkillsRiskValueCosts
Management
55
48
62
50
35
Sales
78
52
41
70
30
Risk and underwriting
46
60
58
55
40
Legal and compliance
32
58
74
38
45
IT
85
76
66
72
58
Operations
18
30
52
66
30
Example values
  • Where to train first
  • Where to act on risks
  • Where to allocate costs
  • Where to bring in or extend AI

What you get

Six documents, each with a name

They can be written into a contract and checked on delivery.

AI usage map

Who uses what, for which activities, with which tools and accounts. And where AI is not yet present.

Maturity index

The 0 to 100 score on the five dimensions, function by function, comparable over time.

Cost baseline

Licences, tokens and infrastructure in one statement per team, including the part that does not reach an invoice today. Where AI has yet to be brought in, an estimate.

Risk register

Unauthorised tools, sensitive data in prompts, gaps in policies. In the format your DPO and compliance team need.

Training plan by profile

Those who do not use it, those who use it badly and those who use it well need three different paths.

Report for management

A written document with the numbers, the use cases to start from and the path we propose.

The data collected stays yours. Repeating the assessment after a few months shows what has changed.

Four weeks, from the first meeting to the report

  1. Week 1

    Setup

    Scope and contacts are defined, and we align with the DPO, HR and IT.

  2. Week 2

    Collection

    The questionnaire opens, system data is collected, interviews take place.

  3. Week 3

    Analysis

    We process the results by function and by profile.

  4. Week 4

    Handover

    We deliver the report and discuss it with management in a workshop.

What we need from you: a sponsor in management, one HR contact and one IT contact, access to consoles and invoices, fifteen minutes per person.

For developers, the same method applied to the code

If the scope is a development group, instead of invoices we read repositories and pipelines. We look at tests, continuous integration, documentation and written rules, and we follow a live working session. The dimensions become adoption, skills, process, code readiness, security and costs. Each team is placed on a five-level scale, from individual use to a governed cycle.

It has to pass compliance first

Three written guarantees, each with a document we hand over.

Anonymity and GDPR

The questionnaire is anonymous and results are read by function, never below five respondents. The impact assessment is done before sending, with your DPO and HR. We hand over the questionnaire's DPIA and the updated record of processing activities.

Local models

Where the path involves models, they stay in your infrastructure or in a European cloud tenant. We hand over the deployment architecture and the model register, with the data location of each.

AI Act

Risk classification for every system surveyed, transparency obligations, human oversight where the regulation requires it. We hand over the AI systems register and the compliance file.

ISO 27001 · ISO 9001 · AWS Partner, Select Tier Services

What happens next, if you want to continue

Each step has a result of its own, useful even if you stop there.

  1. 4 weeks

    Assessment

    Maturity index, cost baseline and report for management.

  2. 2-3 weeks

    Pilot Build

    One team, one use case and one number. You go from "we spend a lot" to "this costs this much".

  3. 6-8 weeks

    Build

    Costs allocated to teams, models chosen by type of work, budget, usage policy and training by profile.

  4. monthly retainer

    Stewardship

    Cost, quality and risk followed over time, with a monthly report and a quarterly review.

Questions we get

What people ask before signing

It depends on how many people are in scope. We write the figure in the proposal before we start, and it does not change once work is under way.

Contact us

A quote, a particular request, or simply a coffee to get to know each other? Write to us and we will reply in a "nanosecond".

Address
Piazza Maestri del Lavoro 7
20063, Cernusco sul Naviglio (MI)
Italy
Address
Piazza dei Martiri 1
40121, Bologna (BO)
Italy