Our thesis
Vertical AI‑native businesses
The thesis behind every company we spin up: an industry chosen with explicit criteria, a service delivered directly, proprietary data that every customer helps generate, and AI as the operating engine. In this order.
01 / Why vertical
The niche is not a limit. It's the barrier.
In horizontal markets you compete on capital and distribution, against everyone. In verticals you compete on depth: knowledge of the processes, the language of the industry, compliance, relationships. These are barriers no funding round can buy: they are built by being inside the industry.
That is why every startup of ours is born in a niche chosen beforehand, against six explicit criteria: a real, deeply felt problem, a clearly identifiable customer willing to pay, a market that is accessible, fragmented and with high transformation potential (the "purple ocean"), and proprietary data that can be captured.
And that is why the co-founder we look for is not a generic digital professional: they are an expert of the vertical, with years of experience spent inside the industry and a bent for frontier technologies.
02 / The difference that matters
Selling tools to an industry's operators, or being the operator.
A software vendor sells a tool and stops at the customer's door: it does not see how the service is delivered, does not own the data that delivery generates, does not control the relationship.
Our companies do the opposite: they are direct actors in their industry. They deliver the service, own the data and control the relationship with the end customer. Revenue proves the service works. Proprietary data is what remains and grows.
It is also what makes the acquisition more valuable for an industrial partner: they are not buying a technology, nor a static asset. They are buying a consolidated market position, the relationships with end customers and a flow of data they could not produce internally.
03 / The moat
The data generated by the service is the defensive asset.
AI models can be rented and they all look alike: what cannot be rented is the operational dataset of an industry. When a business delivers the service directly, every interaction produces data that used to be neither digitised nor structured, and that no external vendor can see.
There is a second reason, harder to replicate than the first. Every startup of ours sits between the two sides of the same process and records both. Not one side of the transaction, but the way the two sides meet, day after day, across thousands of real cases.
This data cycle is designed from day one, together with the service: what to collect, how to structure it, how to put it to work. With every new customer the product improves exactly where generalist models fail, and the competitive advantage strengthens instead of eroding.
- Veterly sits between veterinarians and pet owners: it records an animal's clinical history and care habits across its whole life, today scattered among practices that don't talk to each other.
- Condeo sits between building administrators and residents: it records the life cycle of the building, from maintenance works to suppliers to assembly decisions.
- Sencare sits between caregivers and families: it records the day-by-day course of home care, which until now left no trace.
- ESGmax sits between companies and their sustainability stakeholders: it records environmental and social performance as continuous measurement, not as a yearly report.
- Kontai sits between companies and their recurring suppliers: it records the real behaviour of recurring spend, contract after contract.
04 / AI‑native
AI is not the icing. It's the engine.
AI‑native does not mean adding an assistant to an existing product. It means designing the service on the assumption that AI is its operating engine: workflows, unit economics and customer experience are built on that premise from day one.
That is how a startup gets to deliver a service with an operational efficiency that did not exist in its industry. The same processes, with costs and response times of a different order of magnitude.
This is where the studio's technology infrastructure does its work: shared components, AI frameworks and data flows that every new startup inherits at birth, and the AX (Agentic Experience) expertise to design experiences centred on AI agents.
The hierarchy, however, does not change: first the industry, then the data, finally the technology. AI amplifies an advantage that is born elsewhere; it does not replace it.
This thesis becomes a startup through the studio's method, and it is designed to be repeated.
From niche selection to the choice of the industrial partner.