When a smart factory
stops estimating
and starts knowing.
The client and its context
Varigrafica Alto Lazio isn't a printing house in the traditional sense. It's a smart factory: producing more than 22 million sheets a day, handling around 500 orders daily, with a level of automation many manufacturing companies still look at from afar.
The workflow is already fully digitalised, orders are managed autonomously by the system, every stage of digital production is monitored in real time via machine vision, and the company has invested €10 million in new-generation machinery. The ISO certifications speak for themselves: quality, safety and sustainability aren't slogans, they're daily practice. In short: when Generazione AI arrived, it wasn't about bringing digital into the company. Digital was already there.
Where we started
The problem
And yet, inside this highly advanced environment, there was a contradiction hard to ignore.
The entire digital printing side was under control: numbers, times, performance. But the Offset department, the one with the large rotaries, the most complex jobs, often the most important margins, was still running in a traditional way. Times were logged by hand. Costs were estimated, not measured. And at the end of a job, no one knew for certain whether it had earned or lost money, and how much.
The result? With 500 orders a day to manage, without reliable data it was impossible to anticipate machine downtime, optimise workloads or understand where margin was evaporating.
What we built
Our approach
The goal was clear: equip Varigrafica with a control tower built on Plaintic and the Private Industrial Intelligence framework, a central system capable of collecting, connecting and making readable all production data, in real time.
Not a new ERP. Not a statistics dashboard. A governance tool: something that would let management make decisions based on what is actually happening, not on what is presumed to be happening.
The project developed across four main areas. Master data and structure: complete records of machines and people, intelligent stock management, a physical warehouse map and six cost centres with their hourly cost, all calculated comprehensively, no estimates. Physical and environmental traceability: every pallet, every load now moves under the radar thanks to RFID technology and location sensors with ten-centimetre precision. The environmental sensors introduced an important novelty: paper enters production only when temperature and humidity conditions are optimal. Production under control: every job is tracked hour by hour, with working times, hours worked and waste logged and analysable by root cause. Real-time accounting: for the first time, the real cost of every job is visible during production, not only after the fact, and is automatically compared with the quote.
The team, a front-end developer, a back-end developer acting as project manager, a data scientist and a Trainee Digital Academy coordinator, worked for 12 weeks side by side with Varigrafica's management and administrative leads.
The turning point
Finally we can stop estimating. Now we can know.
The 12 technical weeks were demanding, but the real challenge was different: accompanying an organisation used to deciding by experience and intuition towards a way of working based on data. It's not an automatic transition. It requires dialogue, patience, and above all the tool proving worthy of the trust it's asking for.
It happened in front of a screen. The system had just flagged a margin deviation of -2.39% on a job in progress. It wasn't a catastrophic alarm. It was a precise number, on a specific job, in that moment. And management understood, very concretely, that this was not a reporting system. It was a tool to govern complexity.
From that moment on, the Offset department stopped being a black box. And Varigrafica started building something worth more than any technology: a data culture, the awareness that every minute of processing, every wasted sheet, every machine cycle has a measurable value, and that this measurement is the starting point of any serious decision.
The results
What changed concretely
On the operational front, the Offset department is no longer opaque: processing times, energy consumption, hours per job are now real data, available in real time. The physical warehouse and the "on paper" warehouse finally match. And paper enters the machine only when conditions are right to produce quality.
On the economic front, the system can already flag margin deviations above 2% while the job is still in progress, not afterwards. Quotes, previously built on approximate estimates, can now rely on verified data.
On the quality front, complete traceability and the root-cause waste register have created a continuous improvement channel that simply didn't exist before.
Now the data tells us where we're losing. Before we didn't even know.
What changed isn't only the tool, but the way Varigrafica governs its industrial complexity. An opaque department became a readable system, and data stopped being material for after-the-fact accounting to become a daily decision lever.
What comes next
Turning industrial complexity into something you can see, understand and govern.
Three extensions are already in development: predictive maintenance (knowing when a machine is about to have a problem, before it happens), shift optimisation based on energy costs by time band, and integration with an automated warehouse management system.
The direction is one: more intelligence, less complexity to manage manually, all connected in a single ecosystem.
This experience taught us something important. Working with a company like Varigrafica, already technologically mature, already connected, already advanced, means facing a different challenge from what people usually imagine when talking about digital transformation. It's not about bringing digital. It's about making sure all the intelligence already present in the data finally becomes readable and usable. Turning industrial complexity into something you can see, understand and govern. That's exactly the kind of work we love to do.
It's not about bringing digital. It's about making intelligence finally readable.
Tell us your idea.
Share your vision with us
This is the space where we meet to truly get to know each other, focus on your needs and discover how AI can bring value sustainably.
