Vibeslist:Learning to Read a City

At nineteen, before Google Street View reached Italy, Leonardo started Taranto 360 with two friends.

They carried a homemade photography rig through Taranto, Leonardo's hometown in southern Italy. They took photographs at consistent angles, stitched them into panoramas, and built a way for people to move through the city's streets online. Visitors could explore Taranto and enter local businesses from their computers.

Leonardo wrote the software, spoke with merchants, explored payments, and tried to attract visitors. Without knowing it, he was already learning the generalist reality of building a company.

Taranto 360 lasted only six months. The technology worked, but the project had not started with a problem urgent enough to sustain a business.

The project ended. The dream did not.

Leonardo remained fascinated by the hidden systems beneath everyday life. As a child, after struggling in Italian class, he was given one last chance to improve his grade. His teacher asked him to read a four-page story and summarize it in an hour. She later called his response one of the best summaries she had read.

Leonardo did not understand what had been special about it. Some parts of the story had carried weight. Others had not. He kept the signal and reconstructed the whole around it.

At university, dense algorithm textbooks created the same frustration. Leonardo responded by building an interactive visualizer that let people watch algorithms work step by step. He was not interested in complexity for its own sake. He wanted to reveal the system underneath it.

That became a pattern: study something difficult, find what matters, and build a clearer way through it.

Open source became Leonardo's education in both product and storytelling.

He built Tiramisu as a playful JavaScript experiment at a Python conference. Later, it helped him secure a job in London because an interviewer noticed something unusual in the work: a young developer had cared enough to write tests.

He built Bombolone, a content-management system that he repeatedly reconstructed with different technologies. By keeping the problem steady and changing the machinery underneath, he learned that many systems can produce the same visible result, but their tradeoffs shape everything that follows.

After moving to San Francisco, Leonardo volunteered to give short talks at technical meetups. He wanted to practice English and learn how to explain what he was building. At one of those meetups, an engineer recognized that Bombolone used the exact combination of technologies his team needed. That encounter led Leonardo to Twitter.

At Twitter, he built ng-tasty, a simpler way for developers to work with large data tables. The project spread inside the company and eventually brought Leonardo onstage at ng-conf.

The lesson stayed with him: building and explaining are not separate crafts. A product must work, but the value of the work must also become clear to the person using it.

The hardest systems should feel simple by the time they reach someone else's hands.

Around the same period, Leonardo began a project called OpenTaste.

It organized thousands of recipes and ingredients through search, ranking, and taxonomy. Leonardo wanted to help someone find the right recipe for their preferences and for what was already inside their refrigerator.

But food could not be reduced to ingredients alone. A system needed to understand the distinctions that make something authentic, distinctive, or right for a particular person.

Leonardo found Luca Sessa through his food blog and wrote to him on Messenger. Luca was a food and wine critic who had earned credibility both online and in person. He understood the nearly imperceptible shades of taste that separate a familiar dish from an authentic one, or a popular place from an original one.

Luca described those distinctions. Leonardo tried to make them legible in code.

The two discovered a shared ambition: to use technology to make valuable knowledge accessible without removing the human judgment that gave it meaning.

Around the same time, Leonardo met Giovanni Puntil in London. They connected through a shared enthusiasm for JavaScript, which was still unusual then. Giovanni later contributed to OpenTaste, beginning a working relationship that would continue for more than a decade.

OpenTaste did not become a lasting company. But it gave Leonardo a harder version of his original question:

Can human judgment about a messy world be represented with enough rigor to help another person choose?

In 2015, Leonardo left Twitter to co-found Plan.

The productivity product reached roughly 35,000 monthly users. From the outside, that looked like momentum. Inside the company, customers were repeatedly describing the problems they needed solved while the team concentrated too heavily on technical architecture and the needs of an imagined future user.

Leonardo was building like an engineer when the company needed him to listen like a founder.

Plan did not become the business he hoped it would become.

The experience changed him. He learned that technical quality cannot rescue a product that is solving the wrong problem. Customers often reveal what matters through their behavior before a company knows how to name it. A founder must notice those signals, question their own assumptions, and be willing to cut work they once considered essential.

That lesson would later shape every important turn at Vibeslist.

Leonardo carried those lessons into six years at Coinbase.

One of his first major projects began with a homepage that could take roughly eight seconds to become useful on a constrained connection. He reduced that time to under two seconds and earned broader responsibility for Coinbase's public web experience.

Then he found a deeper problem. Teams were making decisions using analytics events that could be duplicated, missing, or modeled differently across products. Leonardo followed the data from the browser through the APIs and into the systems used by marketing, experimentation, and machine learning.

He replaced layers of complexity with a smaller analytics library that was adopted across Coinbase's web and mobile products. Later, he helped create a shared language for performance across the company, giving teams a consistent way to identify problems and decide what mattered most.

In his final year, he created OnchainKit to make building on Coinbase's Base network feel simple. Developers could create onchain experiences without first reconstructing all the machinery underneath them.

The projects looked different, but Leonardo was still doing the same work: finding signal inside complexity and turning it into infrastructure people could use.

Coinbase taught him that information has no value until people can trust it enough to make a decision.

Whenever Leonardo traveled, the first question kept returning.

He could spend more than twenty hours researching Tokyo, Singapore, Paris, Rome, or London and still feel that he had barely discovered its living soul. The monuments and famous restaurants were easy to find. The city itself remained hidden.

The most interesting information was fragmented across venue websites, newsletters, social posts, images, menus, PDFs, and local calendars. Events appeared, moved, sold out, and disappeared. Sources contradicted one another. Search rewarded what was already visible, not necessarily what was current or meaningful.

When Leonardo left Coinbase in September 2024, he intended to take time off. Two days later, he decided to build Vibeslist.

The first version was a consumer product built around curated lists. Leonardo brought the idea to Luca, and they began comparing their perspectives before the business or product was fully defined. Luca's knowledge of food, hospitality, and local culture challenged the system to preserve originality and authenticity. Their conversations turned Leonardo's initial idea into a shared vision.

The team researched, built, and watched how people behaved. Then it confronted the limits of its first approach. A city changes too quickly for manual curation, and travelers need help only occasionally. The people trying to understand the city every day were inside hotels.

During an early conversation with the concierge team at Hotel de la Ville in Rome, the need became concrete. Even professionals with deep local knowledge could not continuously monitor everything happening around them. They needed a reliable way to see the city's changing events.

That moment did not create the dream. It showed the team where the dream could begin.

Leonardo asked Giovanni to become Vibeslist's CTO. Giovanni said yes within twenty-four hours. He was not betting on a polished pitch. He was betting on thirteen years of watching Leonardo keep moving through difficult problems until he found a way forward.

Giovanni brought the craft required to turn complex infrastructure into a precise, intuitive product. His experience building high-stakes digital experiences and mentoring engineers complemented Leonardo's work in data, taxonomy, and systems.

Jessica Zizzamia joined before her role had a name. For years, she had challenged Leonardo's writing and helped him see weaknesses in work that felt finished. Her training as an opera singer had taught her to hold an entire performance in mind while noticing when a single word, note, or gesture weakened the whole. At Vibeslist, that discipline became part of the company's standard for product quality.

Luca brought cultural judgment, hotel relationships, and a definition of taste grounded in originality and authenticity. He had helped Leonardo encode food knowledge years earlier. Now he helped the team understand what hospitality professionals and their guests actually needed from a city.

The mission had started with Leonardo's lifelong question, but it could not remain his alone. Luca, Giovanni, and Jessica joined before the path was certain. Each brought a different way of seeing what Leonardo could not see by himself.

Together, they became the founding team of Vibeslist.

Vibeslist is building Events Intelligence for luxury hotels, starting with concierge and guest-relations teams who need to read their city every day.

The immediate mission is practical. Help a hotel professional answer a difficult guest question with confidence.

The system continuously discovers city information, connects it to its sources, and structures it around time, place, status, audience, taste, and change. It must understand not only what exists, but what is still true, what is distinctive, and who it could matter to.

When a hotel team asks a difficult question, the answer exposes the strength or weakness of the entire system. A weak result may begin with a missing source, information trapped inside an image, an incomplete taxonomy, a poor connection between two ideas, or a retrieval system that misunderstood the intent.

The Vibeslist team studies those failures from beginning to end. It does not repair only the individual answer. It looks for an improvement that can strengthen thousands of related answers in the future.

This is why hotels are more than a market for Vibeslist. They are the first demanding environment in which Events Intelligence must earn trust. A concierge puts personal judgment and the hotel's reputation behind a recommendation. The information must be current, accurate, distinctive, and relevant enough to share with a guest.

Today, Vibeslist gives expert human agents a wider and more current view of their city. The larger ambition is to build the Events Taste Layer that enterprises and AI agents can use to understand what is happening in the real world, what is distinctive, and who it matters to.

Leonardo does not believe technology should send everyone to the same places.

When a system cannot understand context or taste, abundance collapses into the obvious. The most visible places receive more attention. The rest of the city becomes harder to see, regardless of how meaningful it might be to the right person.

Vibeslist exists to create another possibility.

A place does not need to matter to everyone. It needs to become visible to the person who will value it deeply.

If Vibeslist succeeds, hotel professionals will be able to reveal more of their cities to their guests. Enterprises will be able to create experiences grounded in what is actually happening. AI agents will be able to act with context and taste, not simply repeat what is easiest to find.

Every person, and every agent acting on their behalf, will be one step closer to the freedom to experience a city through their own taste.

Because the soul of a city should not belong only to those who already know where to look.

Somewhere, a dinner is being planned for one night. A gallery is preparing to open its doors. A neighborhood tradition is about to begin again.

The city has already changed.

Vibeslist begins reading it again.