How can a large industrial company bring AI into its processes without stalling at the pilot stage? In this factoryvoice article we tell the story of Fidia Farmaceutici, which chose Skillvue, a Milan-based startup specialising in AI-powered skills assessment, signing a three-year partnership. Gianluca Magnani, Global P&C Officer at Fidia, explains the method and the results: around 400 people assessed in one month, screening times cut from weeks to a few days, and an approach that starts from people development, not just recruitment.
Negli ambienti industriali e corporate, il concetto di Open Innovation viene spesso citato nei report annuali o confinato in progetti pilota estemporanei, pensati più per fare marketing che per trasformare concretamente i processi. Ma cosa succede quando un’azienda farmaceutica globale decide di investire davvero su una startup ad alto potenziale, scegliendo un orizzonte temporale triennale e integrando l’Intelligenza Artificiale nei propri processi chiave?
This is the path taken by Fidia Farmaceutici, a multinational that chose to team up with an innovative startup specialising in AI-driven skills analysis and screening, signing a three-year collaboration.
For our factoryvoice column, we interviewed Gianluca Magnani, Global P&C Officer at Fidia, to understand the origins of this choice, the impact of AI on corporate efficiency, and the key role that large industrial companies can and must play in supporting the innovation ecosystem.
The temptation of “playing it safe” vs the value of calculated risk
When a large company decides to digitalise or innovate a critical process, the most common and reassuring reaction is to turn to big corporates or long-standing suppliers with decades of history behind them. Choosing a startup means accepting a degree of uncertainty and risk. Yet it is precisely within that margin that the real drive for innovation lies.
“As a company that innovates, understanding who is driving innovation in the market is part of our DNA. Valuing people’s ideas translates precisely into stimulating innovative capacity. If this is a company value, it’s clear that I also look for it outside. Finding partners who want to innovate together with us is essential.”
Skillvue, founded in Milan in 2021, closed a €5.5 million Seed round in 2025. According to the HR Innovation Practice Observatory at Politecnico di Milano, in 2024 only 35% of Italian companies managed to identify the “hidden talents” within their own organisation.
Fidia and Skillvue – a Milan-based HR tech startup – met in a university setting, at the innovation labs of Politecnico di Milano. Beyond the effectiveness of the algorithm, what made the difference was the startup’s methodological rigour in validating the tool and in tackling cognitive biases in assessment processes.
“In recruitment, one of the hardest things is overcoming our own biases. When I asked how they had trained the AI to avoid them, I really liked the answer: we can’t avoid them entirely, but we reduce our own. When we receive the screened applications, the manager won’t know the person’s age, gender or educational qualifications, making the assessment as neutral as possible.”
Simplify and streamline: AI at the service of business agility
In today’s industrial landscape, Artificial Intelligence has become an indispensable tool for simplifying work and compressing timelines, turning repetitive or highly complex activities into smooth, effective processes.
In Fidia’s case, adopting AI made it possible to drastically cut execution times, enabling a level of data analysis that would otherwise have been out of reach:
“When you run a traditional screening across a complex population, a whole year has already gone by between the start and the end of the assessment. But at today’s pace, that assessment is already out of date. If you consider that with AI we assessed around 400 people in just one month, how could you ever achieve that with any other tool?”
It’s not just about speeding up analysis, but about generating a real economic and organisational impact on processes:
“If I compress the screening phase from two or three weeks to three or four days, that means saving thousands of euros and getting people into the field faster. On top of that, we don’t bear the inefficiency cost of positions left vacant for a long time. AI lets us make the most of an incredible wealth of data that we already have in the company but often don’t know how to use well.”
Not a “disposable” test, but a strategic partnership
Many startups suffer from Proof of Concept (PoC) syndrome: one-off pilot projects that end up going nowhere. Fidia’s choice stands out precisely for its time horizon: a 3-year contract that turns a supplier relationship into a true collaboration between peers. Fidia acts as a “training ground” and real-world lab, making the complexity of its business processes available to refine and scale the solution.
“Rather than ‘adopting a startup’, it’s about working together as co-protagonists of innovation. I’d like to develop this partnership using our company as a live test bed, with an agile approach: there’s an idea, we try a small development; if it works, we build on it and scale it up.”
From recruiting to people growth: the tangible value
The pilot project started with a highly relevant scope for the pharma sector: Medical Representatives and Account Managers, who make up around a third of the company’s workforce.
Rather than limiting the use of AI to screening during recruitment, Fidia turned the perspective around, starting from the development and mapping of internal skills:
- Unlocking hidden potential: The AI analysis made it possible to identify untapped aptitudes. “Maybe someone isn’t at their best in sales, but it turns out they have fantastic potential for marketing. If you don’t know that, you miss out on an incredible pipeline.”
- Fighting obsolescence and social responsibility: AI helps understand how roles will evolve over the next 2–3 years. “We need to put people in a position to always be ‘employable’. Preventing them from doing only repetitive, low-value tasks that make them obsolete: that is true corporate social responsibility.”
- Onboarding and tailored development paths: The metrics obtained are reused in recruiting. “When a person joins, they already have a development plan ready to work on together with their manager. Especially for younger generations, being able to say straight away ‘here’s what we’ll do for your growth’ makes a huge difference.”
A call to responsibility for the industrial ecosystem
L’esperienza di Fidia dimostra che la digitalizzazione e l’adozione dell’Intelligenza Artificiale non servono a sostituire il fattore umano, ma a liberarlo dalle attività farraginose per concentrarlo sul valore strategico e sulla crescita.
At the same time, bringing in high-potential startups allows established companies to stay agile and keep pace with the times. Startups bring technological freshness and speed of vision; established companies offer complexity, scale and economic sustainability.
Scommettere sulle startup del territorio e adottare strumenti AI per semplificare i processi industriali non è una scelta di facciata, ma una decisione strategica di sopravvivenza, competitività ed efficienza per l’intero ecosistema. Il progetto con Skillvue è stato il primo passo per Fidia in ambito HR, ma la strada per un’Open Innovation concreta e guidata dai dati è ormai tracciata.
Frequently Asked Questions (FAQ)
What is Open Innovation?
It is a model in which a company innovates by collaborating with external players such as startups, universities and research centres, rather than relying solely on in-house R&D. It works when the collaboration becomes part of processes and has a medium-term horizon, like the three-year partnership between Fidia and Skillvue.
How is AI used in recruitment?
AI analyses applications and conducts asynchronous interviews to assess skills and motivation against structured criteria. This cuts screening times and makes it possible to compare large volumes of profiles consistently, leaving the final decision to the manager.
Can AI eliminate bias in recruiting?
Not entirely, but it can reduce it – for example, by hiding age, gender and educational qualifications from the manager in the first stage of assessment. Model validation and human oversight remain essential.
Does the AI Act apply to AI used in HR?
Sì. Il Regolamento UE 2024/1689 classifica ad alto rischio i sistemi di AI usati per selezione, valutazione e gestione del personale. Questi sistemi sono quindi soggetti a requisiti specifici di trasparenza, qualità dei dati e controllo umano.