AI and EU AI Act advisor for technical companies
AI in your company:
adopt it.
I help technical and manufacturing companies use AI where it saves time, build it into their processes, including with models installed on premises, and prepare the controls and evidence the AI Act asks for. It starts with a fixed-price assessment.
Example evidence register
| Obligation | Evidence | Status |
|---|---|---|
| Art. 4 · AI literacy | certificates and attendance | to verify |
| Art. 50 · transparency | notice and dated configuration | to verify |
| AI system inventory | AI systems register | to verify |
| GDPR · impact assessment | Controller’s DPIA, with DPO advice where applicable | to verify |
| Data usage rules | guidelines shared | to verify |
Illustrative example: controls and records depend on the systems and applicable obligations.
Current work: AI adoption in an industrial company, an on-premises model, six applications in production, an AI Act method, monitoring visibility in AI assistants, and online courses with a student area. Clients remain anonymous.
The numbers
In one year, the share of Italian companies using AI doubled.
16.4%
of Italian companies with 10+ employees used AI in 2025 (up from 8.2% in 2024).
15.7%
of SMEs, against 53.1% of large enterprises.
Large companies are racing ahead and SMEs are falling behind: that is where a practical method makes the difference.
2023
2024
2025
Source: ISTAT, Imprese e ICT, anno 2025 (15 dicembre 2025)
Source: Eurostat, Use of artificial intelligence in enterprises (dati 2025, estratti a dicembre 2025)
Why companies hold back
What's missing: skills, clear rules and safeguards for data.
Skills, clear rules and data protection are among the barriers reported by businesses that considered AI but did not adopt it. I work with your teams on these issues.
-
Lack of skills 58.6%
-
Unclear rules 47.3%
-
Data unavailable or of poor quality 45.2%
-
Privacy and data protection 43.2%
-
High costs 43.0%
-
Ethical considerations 25.7%
Adoption, development and governance in one programme
Adoption, development and governance meet in the same processes. I define controls, training and documentation around the use of AI and the applicable obligations.
Adopting AI
Your people already use AI, each in their own way. I help them use it more thoughtfully, with guidelines developed together and training tailored to their roles.
How to adopt it 02Building
Agents, integrations and models that stay in-house, with automated tests, logs and human approval. Systems designed to go into production, tested before every release.
What I build 03Getting ready for the AI Act
Inventory of systems, role, risk, obligations and evidence. A technical method, with your lawyer where needed.
The methodThe method
Every obligation becomes a control. Every control produces evidence.
-
01
Obligation
Art. 4: AI literacy of staff
-
02
Control
Role-based learning path and usage guidelines
-
03
Evidence
Certificates of attendance and attendance register
-
04
Status
Done, with a review date
AI Act
Some obligations already apply today
-
1 Aug 2024
Entry into force
-
2 Feb 2025
Prohibitions and AI literacy
-
2 Aug 2025
General-purpose AI models
-
2 Aug 2026
Transparency and general application
-
2 Dec 2026
New prohibitions and content marking
-
2 Dec 2027
High risk, Annex III
-
2 Aug 2028
High-risk Annex I systems; AI requirements for machinery
From first step to ongoing support
-
01
Assessment
A fixed-price assessment of where AI could add value, the systems you already use and the AI Act requirements to consider.
-
02
Training
Role-specific training and records of the programme delivered, with content and availability to be agreed. Online course in preparation.
-
03
Build-Lab
A project developed with your teams: AI agents and operating rules tailored to your processes.
-
04
Ongoing support
Ongoing advisory: a review every six months, and whenever the rules change or a new system arrives.
The assessment in practice
- What you get
- A map of the AI tools already in use, the opportunities worth pursuing ranked by impact and feasibility, the AI Act points to check, and an action plan for the first three months.
- Who is involved
- Technical management, IT and two or three department leads.
- Price
- Fixed, agreed before we start. Build-Lab and ongoing support are quoted separately.
Who is behind it
I'm Valerio Nicolai, based in Italy, with a degree in computer engineering (industrial automation). I have spent years in regulated sectors, from public transport to digital projects for public bodies. Today I advise an industrial company on AI adoption and build platforms that run in production.
My backgroundThe magazine
ReFocused
The editorial project covering applied AI, the AI Act and governance. Explore the available articles and the local public transport archive.
Read the articlesTraining
Online courses
Currently two free video courses on local public transport, with a members' area. An AI literacy programme for companies is in preparation.
Go to trainingLet's start with how you use AI today
Tell me which tools your people use and what you want to achieve. I will reply personally, with the first step I would take and the AI Act points to check in your case.
Get in touch