Bringing artificial intelligence into a business starts with a question: what problem do we want to solve? The next step is to understand what data is available, what expertise is needed and how to assess whether the proposed solution delivers a tangible benefit. Working with researchers can help businesses evaluate the technology’s potential and limitations throughout this process.
We discuss this with Dr Silvia Boi, Innovation Manager at Fondazione FAIR. At Borsa della Ricerca 2026, she met with companies to discuss the opportunities offered by the Foundation’s scientific network and the SINTESI and AI4Nature innovation hubs. These discussions offered an opportunity to explore how a business need can lead to collaboration with researchers, what it takes to launch a pilot project and which questions companies should ask before investing in AI.
When a company contacts you about “bringing artificial intelligence into the business”, what is the first question you ask, and how do you move from that broad intention to identifying a specific problem?
“The first question we usually ask is: ‘What problem do you want to solve, and what outcome would you like to achieve?’ Companies often start with an interest in the technology, but our job is to help them identify a concrete need: reducing waste, improving a service, making better use of available information or making a repetitive task less burdensome.”
“We then look more closely at the process involved, the current difficulties and how improvement could be measured. This discussion helps turn a broad intention into a use case for which we can bring in the appropriate scientific expertise. This is also the approach we aim to support through the SINTESI and AI4Nature innovation hubs: connecting research capabilities with business needs.”
How do you determine whether a problem really requires AI, and when would you recommend first addressing work processes or data, or using simpler tools?
“We assess whether AI can offer a tangible advantage over the alternatives. It can be useful, for example, when you need to identify patterns in large volumes of data, analyse images and text, or make predictions about complex phenomena. But the choice always depends on the context and the quality of the available information.”
“If the problem stems from unclear procedures, fragmented data or systems that cannot communicate with one another, these issues often need to be addressed first. In other cases, rules-based automation may be enough. Recognising that AI is not the most suitable solution is itself a useful outcome of discussions with researchers: it helps a company allocate its resources and investments more effectively.”
At Borsa della Ricerca, you met companies interested in FAIR’s expertise and opportunities. Could you describe a specific request that emerged from those discussions, without necessarily naming the company? What further investigation is needed to establish whether it could become a project?
“A range of needs emerged at Borsa della Ricerca. Some companies wanted to understand how to integrate AI into their applications and which areas of expertise within the FAIR network could help them. Others expressed an interest in testing the services offered by the SINTESI and AI4Nature hubs or taking part in training activities to develop the skills needed to assess and use these technologies. There was also interest in building new research and innovation partnerships. In healthcare and oncology, some companies also expressed a willingness to share data to support the development and validation of AI solutions.”
“Turning these initial contacts into opportunities for collaboration requires a deeper understanding of the need, the identification of suitable expertise and a first objective whose achievement can be verified. Depending on the case, the next step might be a technical discussion with researchers, a training programme or the design of a pilot project, with clear requirements for data, resources and responsibilities.”
What should a company bring to its first meeting with you? Are a problem and an objective enough, or does it already need usable data, in-house expertise and a budget?
“For the first meeting, a concrete problem and a willingness to explain it clearly are enough. We don’t expect a company to arrive with a fully developed project. It helps to involve someone with first-hand knowledge of the process who can explain where difficulties arise, how they are currently addressed and what kind of improvement would be valuable.”
“Data, in-house expertise and budget are matters to explore together. We need to understand what information exists, whether it can be used and what resources a potential pilot project would require. It is also very important to identify a company contact who can stay involved throughout the process: collaboration requires an ongoing exchange between those who understand the problem and those who develop the solution.”
When is it useful to involve a research team rather than choose a solution already available on the market? How do you identify the right expertise within the FAIR network, and what is your role in that process?
“An existing solution may be the right choice when it meets a clearly defined need and its performance can be verified in the company’s operating environment. Involving researchers becomes particularly useful when the problem presents new challenges, requires a combination of different areas of expertise or raises unresolved questions about the technology’s performance, reliability or adaptability.”
“Within the FAIR network, including through the SINTESI and AI4Nature hubs, we start by analysing the need and matching it with the available scientific specialisms and infrastructure. My role as Innovation Manager is to facilitate this process: helping the company articulate its need, identifying the relevant researchers and supporting the development of a plan with shared objectives and assessment criteria. The value lies in creating the conditions for those discussions to become a concrete collaboration.”
What can researchers learn from engaging with businesses, and how do you gather these insights within the SINTESI and AI4Nature hubs?
“Engaging with businesses helps researchers understand problems, constraints and working conditions that do not always emerge in the laboratory. A company brings knowledge of its processes, the experience of its people and, in some cases, data against which solutions can be tested. This contribution can generate new research questions and help establish which results are genuinely useful and which aspects require further development.”
“Within the SINTESI and AI4Nature hubs, we work to encourage this exchange through meetings with businesses, the identification of their needs and discussions with research teams. The aim is to build collaborations in which researchers and businesses work together to define the problem and assess the results: companies gain access to new expertise, while researchers can tackle needs and situations arising in practice.”



