![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 22 22" xmlns="http://www.w3.org/2000/svg"><path d="M 0 0 L 0 19.127 L 8.709 22 L 8.709 5.347 C 8.709 0 0 0 0 0 Z M 5.532 19.142 L 3.177 18.371 L 3.177 2.631 L 3.592 2.759 L 4.224 2.953 C 5 3.192 5.532 3.934 5.532 4.776 Z M 18.741 6.199 L 18.741 0 C 18.741 0 10.031 0 10.031 5.347 L 10.031 22 L 18.74 19.127 L 18.74 8.753 L 14.529 10.058 L 14.529 13.606 L 15.564 13.306 L 15.564 18.371 L 13.208 19.142 L 13.208 4.776 C 13.208 3.934 13.741 3.192 14.517 2.953 L 15.564 2.631 L 15.564 7.251 L 18.74 6.199 Z" fill="rgb(0, 0, 0)" height="22px" id="iRPdBku8W" transform="translate(1.63 0)" width="18.740741374796762px"/></svg>)
Nexhub is a non-profit platform that makes innovation resources and entrepreneurial guidance accessible to aspiring founders across Italy, especially in the less-served regions. It maps both institutional players, such as incubators and accelerators, and grassroots organisations like local associations and community groups, so people can see where resources exist and whether they are worth reaching. At its core is AI-supported matchmaking between startups and incubators, based on budget and stage of development.
The engagement covered product and UX end to end: mapping the matchmaking logic between founders and incubators, and designing the platform itself. The platform includes a CMS that lets each organisation self-manage its profile, add events and keep its information up to date.
Matchmaking built for founders starting from zero
Nexhub's real challenge was serving the youngest, lowest-budget startups, usually the hardest to grow. The platform matches founders with incubators and partner companies based on growth stage and trajectory fit, not just sector, powered by an integrated AI engine that surfaces the most relevant matches right after login. From there, founders can review the detail, compare, and choose before moving forward.
A CMS built for circuit quality, not just content
For operators, including incubators, associations, and even software houses managing multiple startups at once, I designed a CMS with dedicated onboarding to keep data quality high at the source. An LLM auto-extracts tags from submitted content, feeding directly into the matchmaking engine. The validation threshold, around 75% on wireframes before aesthetic refinement, wasn't a methodology footnote. It was chosen specifically to keep the MVP scalable and feedback loops fast.
