AI feasibility
Can it be done, and is it worth it? Is the data there, and can it be used? What does it cost to prepare it and to run the service over time? The checks to make before any technical choice.
These pages are for those who run a small or medium-sized business and are considering bringing AI into an application or a process. You do not need technical skills to read them: they are there to help you know what to ask, of a supplier or of your own technical staff, before committing time and money.
In 2025, 15.7% of Italian small and medium-sized businesses with at least 10 employees used at least one AI technology, against 53.1% of large ones; among those that considered it without adopting it, the most cited obstacle is a lack of skills (Istat, Imprese e ICT 2025 (opens in a new tab), in Italian).
I have been developing software for nearly thirty years. With AI I do two things: I assess with you whether the project is sustainable (available data, costs, risks, ongoing operation) and, if it is, I build it. I work on the individual project, not on general strategies.
I have been studying machine learning and language models since around 2020: that is the foundation I work from. My direct experience on real projects is much more recent, and the project described in these pages is still under development and testing. That is why every assessment starts from tests on your own case, not from promises.
Each page answers a question worth asking before you start.
Can it be done, and is it worth it? Is the data there, and can it be used? What does it cost to prepare it and to run the service over time? The checks to make before any technical choice.
A concrete example. Querying business data in natural language, with answers that must match the management system: what it really involves. Project under development and testing.
What does it cost, and how long will it last? The cost of the whole service, not just the model; what happens when a provider withdraws a product; when it pays to wait.
What am I risking? Which data the system can reach, what it can do on its own, who oversees it when it gets things wrong.
Do you develop software and are you interested in AI as a working tool? How I use coding agents