I design and build software for complex industrial systems, with a strong focus on CNC machinery, HMI platforms, CAD/CAM workflows, legacy modernization and applied artificial intelligence.
My background is mainly in long-lived C++/C#/Python codebases, production software, machine-oriented logic, computational geometry, diagnostics, automation and human-machine interaction.
Most of my professional work happens inside private enterprise repositories and NDA-bound environments.
This profile contains the public-facing subset of my engineering work, research experiments and independent projects.
My work sits at the intersection of industrial software, machine automation and applied AI.
I am especially interested in systems that must remain understandable, maintainable and reliable over time: software that talks to machines, supports operators, assists engineering workflows and has to survive real production constraints.
Main areas:
- CNC and CAD/CAM software
- HMI design for industrial machinery
- Legacy C++/C# codebase modernization
- Multi-axis kinematics and machine-oriented computational geometry
- Internal tooling, QA automation and diagnostics
- Applied AI, ML and agent-based workflows for industrial use cases
- Research-oriented simulation and experimental software
Ground Control LEO satellite networks, policy decisions, falsification-driven simulation |
Spectral Generalization Frequency-domain detection of synthetic and AI-generated imagery |
StoryHash Multi-modal fingerprinting of video content for robust retrieval |
More on the research
Ground Control is a discrete-time simulation framework for studying policy decisions in LEO satellite network scenarios.
The project focuses on falsification-driven analysis rather than optimistic simulation: routing policies, DTN behavior, gateway placement, inter-satellite links, SLA trade-offs and topology constraints are tested through reproducible experiments.
Current public material includes:
- a simulation engine
- documented experiment batches
- OneWeb TLE-based validation through SGP4
- research notes and whitepaper material
- policy-level analysis for LEO constellation behavior
Spectral Generalization explores frequency-domain methods for detecting synthetic and AI-generated imagery.
The project investigates whether spectral artifacts can generalize across generation models, datasets and image domains.
StoryHash is an experimental project about multi-modal fingerprinting of video content for robust video retrieval.
The goal is to represent video sequences through a single compact vector that fuses visual, temporal and structural signals, useful for similarity analysis, authenticity research and content-level comparison.
paperdown Markdown to PDF on the command line, with real typography instead of browser defaults. |
ProdKit An open-source command-line engine that reads a repository and reports what is missing before it can carry real users, judged as the kind of product it is meant to be. |
envlockup An encrypted .env you can commit: everyone decrypts it with their own SSH key, so people can be added or removed without sharing a secret. |
QuestEd A platform for courses, exercises and assessments for teaching programming, built with Django. |
qDBManager A lightweight ORM / entity manager for the Qt framework. |
PyImgDuplicateFinder A Python utility to detect and move duplicate images using HaarPSI, a perceptual similarity index. |
Asteroids A C++ implementation of the classic arcade game, developed as an exploration of game loops, rendering and basic engine structure. |
SnapBiz Platform for managing photography businesses |
Scribit.ai Transcription and document translation |
SnapBiz is a full-stack platform for managing photography businesses, including communication flows, client portals, reminders, contracts, internal collaboration and operational dashboards.
Built mainly with Django, Celery, REST APIs and automation-oriented backend workflows.
It is progressively evolving toward a SaaS model with analytics and AI-assisted business workflows.
Scribit.ai is a transcription and document translation platform: recordings become usable text, and Word and PDF files are translated keeping their original layout, with integrity and glossary checks.
Private tools and prototypes covering:
- 5-axis kinematics
- angle computation
- ISO code generation
- post-processing
- machine simulation
- validation utilities
- ML-based diagnostics and pattern analysis
I write about AI, software engineering and the impact of emerging technologies on industrial work.
My focus is not hype, but what happens when AI enters real organizations: legacy systems, production constraints, human workflows, costs, safety, maintainability and decision-making.
I publish mainly on LinkedIn and on Substack, where I collect longer-form essays and research-oriented reflections.
I also mentor students and colleagues on Git workflows, computational geometry, machine learning, game development, CNC/HMI architecture and AI explainability.
The essays are written in Italian.
- Cosa direbbe di noi un contadino del 1750? · October 2026
- “Nessuno mi uccide con l’inganno, non con la forza.” · September 2026
- Ci vuole una città · August 2026
- OpenAI, Hugging Face e l’IA che ha aggirato l’esame · July 2026
- Emma-5, l’IA italiana e il problema di quando il sogno arriva prima del prodotto · June 2026
I tend to work on software where shortcuts become expensive.
My engineering approach is shaped by industrial constraints: hardware limits, safety, latency, cost, backward compatibility, maintainability and long product lifecycles.
I value:
- correctness before cleverness
- explainability over blind optimization
- measurable improvements over theoretical elegance
- simple tools until complexity is justified
- production behavior over demo behavior
- maintainability as a first-class design constraint
I am skeptical of black-box solutions when they are placed inside operational systems without evaluation, observability or failure boundaries.
CAD/CAM · CNC · HMI · Industrial Automation · Machine Simulation · Applied AI · ML Prototyping · Agent Workflows · Software Architecture · Legacy Modernization
I am inventor or co-inventor of multiple software patents related to CNC machinery, HMI workflows, CAD/CAM automation and machine-control logic.
Details are available only where disclosure is permitted.
This profile is a public snapshot of a broader technical path: industrial software engineering, applied AI research, independent experimentation, teaching and product-oriented side projects.
The common thread is building software that has to deal with reality: machines, users, constraints, costs, failures and long-term maintenance.
For technical discussions, research collaboration or professional inquiries, feel free to reach out.




