About us · Varese, since 2021

Production first. Then the lecture hall, by invitation.

We were born in 2021 in Varese as UEBB Artificial Intelligence, with one clear idea: bring Artificial Intelligence where it really matters — production, plants, quality, safety, medicine. Seven people, vertical skills, everything in-house: and when universities call, we bring to the classroom what we have already put into production.

Talk to us University guest lectures

Varese · since 2021
Team of 7
AI-first since day one
NVIDIA · Sophos · Veeam

The team

Seven people, zero handoffs

Vertical skills covering the entire life cycle of a system: from the model to the firmware, from the network to security. The people who design are the people who build, and they stay by your side after the release. And when a project calls for one more speciality, selected external collaborators work actively with us, inside our method.

Founder & CEO

AI architectures, project direction and a direct relationship with every client.

AI Engineer

Models, training and MLOps: from GA.IA to on-premise RAG systems.

Full Stack Developer

Web, APIs and integrations: the software around the models.

Senior Network Engineer

Enterprise networks, firewalls, VPNs and segmentation.

System Engineer

Servers, virtualization, GPUs and on-premise environments.

CyberSec Specialist

Endpoints, EDR/XDR, hardening and incident response.

External collaborators

Selected specialists who work actively on our projects when one more skill is needed — inside our method, to our standards.

The UEBB studio in Varese during a development session, with a workstation at work in the background — The studio · Varese

The UEBB studio interior in Varese at the end of the day, with workstations and the common area — End of day · 19:08

Explore our AI capabilities →

The story

Three chapters, one idea

012021

Born AI-first

When applied AI was still rare among local companies, UEBB was born vertical on Artificial Intelligence: models, data and infrastructure first. The name already said it: UEBB Artificial Intelligence.

02GA.IA · NVIDIA

Putting things into production

GA.IA, the supply chain engine with nine machine learning models, was born this way: real problems, real data, measured metrics. The NVIDIA Inception program brings resources and technologies straight into our projects — and the work speaks loudly enough to get invited into the classroom, at Insubria and LIUC.

032026

The complete stack

Today UEBB covers everything AI needs to actually work in a company: software development, cloud (AWS, Azure, Google Cloud), cybersecurity with Sophos (Silver Partner), backup and data resilience with Veeam (Cloud & Service Provider), UEBB Drive for data that must never leave home, and EU AI Act compliance. Still from Varese, Via M.V. Staurenghi 34.

UEBB at universities

Our work, brought to the classroom

When Insubria and LIUC want to explain applied AI, they invite those who put it into production: UEBB projects and methods have been brought to the classroom by our CEO in four guest lectures, by invitation. The topics, with the original slides.

University of Insubria01 / 04

Giacomo De Benedetto during the lecture on the study for Not Just Analytics at the University of Insubria

Classifying Instagram content with AI

A study carried out for Not Just Analytics: how to automatically classify the content of Instagram profiles by combining computer vision and language analysis.

Computer visionNLPClassification

University of Insubria02 / 04

Giacomo De Benedetto presenting the NVIDIA ecosystem for HPC at the University of Insubria

Infrastructures for Big Data and HPC

18 slides

From Hadoop, Spark and Airflow to HPC architectures: CPU vs GPU, CUDA and CUDA-X, cuDNN, NVIDIA Nsight and Triton Inference Server, distributed training with PyTorch DDP and NCCL, up to Databricks.

HPCNVIDIA CUDAPyTorch DDP

LIUC – Università Cattaneo03 / 04

AI for the Supply Chain

12 slides

Demand forecasting, scheduling with real constraints, warehousing, quality control with computer vision, predictive maintenance, logistics and risk control towers: where AI delivers measurable benefits.

ForecastingSchedulingPredictive maintenance

LIUC – Università Cattaneo04 / 04

Giacomo De Benedetto presenting GA.IA at LIUC – Università Cattaneo

GA.IA: a concrete example

Nine models in a single platform: LSTM for time series (94.5%), Prophet + XGBoost ensemble (96.6%), Isolation Forest, K-Means + UMAP, DistilBERT, Random Forest (>90%). In the demo: forecast errors −18%, on-time delivery +15%, response times −25%.

GA.IA9 ML modelsControl tower

Beyond the classroom

Satellite imagery

Contributions to AI-based satellite image analysis projects, with publications and numerous citations.

Medical imaging

Ongoing medical imaging projects with Artificial Intelligence.

NVIDIA Inception

Members of the NVIDIA program for AI startups: access to resources and technologies we bring straight into projects.

The founder

Giacomo De Benedetto

Founder & CEO

“I founded UEBB Artificial Intelligence to bring AI where it really matters: production planning, plant reliability, in-line quality, safety and medicine.”
From the guest lecture "AI for Supply Chain" at LIUC – Università Cattaneo

He founded UEBB in 2021 and personally follows AI architectures, applied research — satellite imagery with publications and citations, medical imaging — and the EU AI Act regulatory side. By invitation, he brings UEBB projects to the classroom at Insubria and LIUC.

LinkedIn profile ↗

Partners and recognitions