AI Governance & Security Architect

Stefano Noferi

I design the infrastructure that makes AI agents verifiable and governable in healthcare and enterprise settings.

Entrepreneur, software architect and researcher

Artificial IntelligenceCybersecurityHealthcare Systems
27+years of experience 30+R&D projects 1000+professionals trained
Thoughts Let's build together
Eadweard Muybridge's 1878 plate The Horse in Motion: twelve black-and-white photographs in three rows show a horse and jockey galloping against a background of numbered lines. In the last one the horse is standing still. AI Governance

A two-speed world

On 13 October 2026 Anthropic described how it replaced git in its monorepo, which went from 2,500 to more than 30,000 merged changes a day in four months. It follows on from Intelligence is eating software: work now runs on two lanes, an agentic one and a human one, and governing the first becomes the priority of the second. Ink drawing of an Ouroboros, a black-headed serpent biting its own tail to form a ring. The body is pale and speckled, and in the middle of the ring is the Greek inscription «ἓν τὸ πᾶν». Software Development

Intelligence is eating software

 Detail of the cover of the paperback edition of Neal Stephenson's The Diamond Age. In a red corridor streaked with light a girl runs towards a white opening, and below her figure is the title in silver letters. Hardware

Building machines from code the way a cell does

 Two side-by-side charts on a dark blue background, showing changes in percentage points of GDP from the start of each period. On the left, 1995-2000, a light-blue line and a coral line rise together to about +1.2. On the right, 2023-2026, the light-blue line rises to almost +1 while the coral line falls to about -0.6. AI Governance

AI and the rest of the economy

 Printed page in German type from the Berlin Academy's report of November 1859, with the opening of Riemann's memoir on the number of primes less than a given magnitude and, at the bottom, Euler's product formula. AI Governance

Is the world we knew really over?

 Bifurcation diagram of the logistic map on a midnight-blue background: from the left a thin white line splits in two, then into four and eight ever closer branches, until on the right it opens into a dense cloud that shifts from light blue to coral. AI Governance

AI is already building the AI that comes next

 The Codendum symbol: two blue angle brackets around a blue hexagon, linked by lines to three nodes, two teal and one red. Teaching

Codendum, shared generative coding on local infrastructure

 Negative-stain electron micrograph of Staphylococcus aureus bacteriophages. Dozens of particles with a dark oval head and a long, thin tail, scattered across a pale background. AI Governance

A new family of enzymes found by agents reading DNA

 Gerardus Mercator's 1569 world map reassembled from the eighteen original sheets. The Latin title runs across the top, and dozens of compass roses across the oceans radiate networks of straight lines. Sailing

I will keep using Mercator charts because they do well what they were made for

Topics

Healthcare AI48 AI Governance33 Open Source29 Infrastructure28 Cybersecurity26 Software Development23 Hardware17 Teaching7 Music2 Sailing1
214 posts in the archive

Research

Applied research on AI systems architecture, digital health, and distributed infrastructure. Over 30 funded R&D projects in collaboration with universities, hospitals, and industry partners.

Go to Research

Ventures

Products and solutions across AI, Cybersecurity and Digital Health — open, on-premise, EU-compliant (AI Act, NIS2, GDPR).

View all ventures

Projects

Admina

Open-source framework for AI agent governance. A transparent proxy built on 6 pillars.

OISG

Open, Intelligent, Secure, Governed paradigm for autonomous AI systems.

Thoughts

AI agents are becoming autonomous actors inside software systems.

Current software architectures were not designed to control, monitor, or govern autonomous AI behavior.

Without new architectural layers, AI systems will introduce significant risks in security, accountability, and reliability — especially in critical domains.

My work focuses on designing secure and governable AI systems for healthcare and enterprise environments.

Open Source

Active contributor and maintainer of open-source technologies related to AI systems.

Events & Conferences

Participation and talks at conferences, meetups and community days dedicated to open source, Linux and open technologies.

Consortia & Commercial Open Source

Collaborations with consortia and companies for the development and adoption of open-source solutions in enterprise and industrial contexts.

What I'm Working On

AI Agents Governance

Building infrastructure to control and govern autonomous AI agents.

Agent MonitoringGovernance LayersPermission ModelsLifecycle Control

AI for Clinical Systems

Developing AI decision support systems for clinical and nutritional healthcare environments.

Clinical Decision SupportHospital Data IntegrationPatient Monitoring

Secure AI Architectures

Designing security architectures for AI-driven systems.

AI System SecurityML Infrastructure ProtectionThreat Models for AI

Areas

For over 25 years I've been building software, companies and bridges between different worlds — from cybersecurity to digital health, from academic research to industry.

Linux & Open Source

Passion for free software, community contributions and adoption of Linux as main platform.

Telecommunications & Cybersecurity

Experience in networking, telecommunications and cybersecurity.

noze

Products and solutions across AI, Cybersecurity and Digital Health — open, on-premise, EU-compliant (AI Act, NIS2, GDPR).

Research & Digital Health

Digital health research projects: digital twin for rare diseases.

Training & Coaching

Teaching and coaching for companies, universities and ITS Academy.

University & ITS lecturing

1st and 2nd level masters on cybersecurity, intelligence and open source software.

AI Governance

DebugABot and Admina: platforms for AI agent security and governance.

OISG

Open, Intelligent, Secure, Governed paradigm for autonomous AI systems.

About

Stefano Noferi

Tech entrepreneur based in Pisa, founder of noze in 2000. For over two decades I've been working across artificial intelligence, cybersecurity and healthcare systems. I propose the OISG paradigm for autonomous AI systems. Sailor, sommelier and radio amateur.

Go to About

Stefano Noferi ↓ @stefanoferi / stefanòferi /