AI-AUDIT.IT · MILAN, IT · MMXXVIACCEPTING 2 PROJECTS · Q3

Italian AI.Compliant.Competitive.Measurable.Defensible.Competitive.

Audit, development and AI strategy in a single partner. We build intelligent systems compliant with the EU AI Act for Italian companies that want to use AI effectively and safely.

02 / Projects

Every project,
a sub-brand.

The proprietary products born from our software house — not the services we offer to clients. Each project has its own identity, metrics and page.

PRJ-01

MediTrace.ai

ONLINE · MACOS

Research assistant for Italian oncologists. Queries 12 biomedical databases, synthesizes in Italian with verified citations. Local LLM: data stays on the doctor's Mac.

  • 12 biomedical APIs · PubMed, NCT
  • Local LLM · Qwen 2.5 14B
  • AI Act Art. 12 · Annex IV
PRJ-02

BandoBuilder.it

ONLINE · BANDOBUILDER.IT

The AI platform that simplifies access to European and Italian grants and funding for SMEs. Find, analyze and apply in a fraction of the time.

  • 4,200+ grants monitored
  • 89% AI match · explainable
  • Pre-filled documents
PRJ-03

+Human

Private beta

The marketplace that connects AI agents with human operators to complete tasks in the physical world. API-first, on-demand, certified.

  • AI agent orchestration
  • Certified operators
  • API & webhooks
PRJ-04

MiroFish-IT

Open source · MIT

Multi-agent simulation engine with the ICF framework. Synthetic European populations calibrated on institutional data, NUTS-2 granularity.

  • 32,000 agents simulated
  • Eurostat · ISTAT · Banca d'Italia
  • MIT license · open source
PRJ-05

Quieta.me

Private beta · iOS

iOS app that uses RSVP as a reading rhythm to recondition attention and break compulsive scrolling. Two modes: Now for impulse moments, Path for deliberate reading.

  • RSVP · attention reconditioning
  • Two modes · Now / Path
  • Privacy-first · 100% on-device
PRJ-06

Dual Use

On request · Demo

Assistant for dual-use export control (EU Reg. 2021/821): from the Annex I control entry to the «Y»-code declaration, with the rule always cited and updated to 2026.

  • Annex I · 10 categories · 380+ entries
  • CN/CPA pre-screening · hybrid retrieval
  • On-device · traceable references
Every project is born from our AI Software Development serviceSee how we work
03 / Method

Four phases.
Zero surprises.

A repeatable, measurable, transparent process. Each phase produces verifiable deliverables and decisions that stay in writing.

PHASE 01

Discovery

We map the AI landscape: systems, vendors, goals, constraints. Interviews with management and technical teams, process analysis, a signed scope document.

+
PHASE 02

Analysis

Technical audit of architecture, data quality, performance and compliance. Vendor and supplier due diligence. Risk classification under the EU AI Act.

+
PHASE 03

Report & recommendations

Detailed report with intervention priorities, timeline and cost estimate for each recommendation. C-level presentation with an operational call to action.

+
PHASE 04

Execution & review

We support implementation with coaching, review and governance. KPI monitoring dashboard. Model retraining. Continuous updates.

+
04 / EU AI Act

Risk
classification.

The European Artificial Intelligence Regulation introduces three levels of risk. Every AI system in use at your company belongs to one of them.

Level 01 · EU AI ActProhibited
01

Unacceptable risk

Systems banned by the Regulation. Use contrary to the Union's values, no exceptions.

  • Government social scoring
  • Subliminal manipulation
  • Exploitation of vulnerabilities
  • Mass biometric surveillance
Level 02 · EU AI ActRegulated
02

High risk

Subject to strict requirements for documentation, human oversight and periodic audit.

  • Biometric identification
  • Critical infrastructure
  • Credit scoring
  • Recruitment & HR
Level 03 · EU AI ActTransparency
03

Limited risk

Transparency obligations toward the end user. Clear disclosure, no deception.

  • Chatbots and AI assistants
  • Content generation
  • Emotion recognition
  • Biometric categorization
05 / AI Training

Train your team for the AI Act.
And for real work.

Our training combines compliance, productivity and governance. Not generic tool courses: Ai-Audit paths designed for roles, sectors and concrete use cases. We train people, teams and management so that AI is used with method, safety and competitive advantage.

AI literacy compliant with Art. 4 for all staff who use AI systems

Vertical upskilling for functions, professions and specialist teams

Governance and policy to introduce AI into the company in a controlled way

Ai-Audit certified paths with a practical, operational focus

We believe artificial intelligence is the most powerful lever for the competitiveness of Italian companies. But only if implemented with competence, strategic vision and an ethical approach.

AL
Andrea Lorini
Founding Partner · Ai-Audit.it
0
Audits completed
0
Enterprise clients
0%
Avg ROI · 12 months
0%
Client retention
06 / Contact

Let's talk.
First call
free.

Tell us about your project. We reply within 24 hours. No commitment, no conflict of interest.

HELLO@AI-AUDIT.ITMilano