• Orange brings its technical expertise (cybersecurity, telecoms, edge computing) and a potential commercial outlet, which is the best sign of funding for a startup.
• The association’s general manager, Romain Roullois, presents the project and the benefits of the partnership at VivaTech 2026, where the association and Orange are in attendance.
What is your definition of deep tech in 2026?
Romain Roullois. Deep tech is a cross-cutting characteristic of certain companies or projects, which can be summarized as the scientific intensity embedded in a company’s value proposition. In quantum computing, almost all startups are deep tech; in food tech, a food bioproduction company is, but not a delivery service. Specifically, there are four key indicators: R&D spending exceeding 20% of expenses, at least one scientific founder, a direct link to research—that is, a contract or origins in a laboratory—and, finally, an innovation protected by intellectual property. In 2026, the wave of technological sovereignty is heavily focused on this topic: many companies are seeking to align themselves with it, and this lends credibility to their projects in the eyes of investors and partners.
A contract is the best possible signal for funding, ahead of a fundraising round or a grant, because it proves that there is a market need
What is the purpose of the France Deeptech association that you lead?
The association was founded three years ago to bring together players who did not identify with existing intermediary bodies, as their challenges differ so much from the rest of the tech sector. We now have 350 members: two-thirds are startups and scale-ups, ranging from quantum computing to New Space, such as Pasqal, Alice & Bob, The Exploration Company, and one-third are divided among specialized investors, research institutions (CNRS, CEA, Inria, Polytechnique, ENS, etc.), partners (lawyers, banks, intellectual property consultants), and major corporations like Orange. We have two missions: to bring the ecosystem together through events, and to represent it in order to influence institutions, both in France and at the European level.
What are the specific challenges facing these startups?
First, financing their growth, as there is a lack of investors in France and Europe capable of leading funding rounds of 50 to 500 million euros. Next is the exit [the sale of companies]: the exit process remains poorly structured on this side of the Atlantic, even though it is essential for both entrepreneurs and the funds that generate returns and then invest in other startups. We can add to this the challenge of scaling up, which is particularly delicate in this sector. Nevertheless, France is well-positioned and has around 4,000 deep tech startups, ranking second in the rate of company creation behind the United Kingdom and ahead of Germany. This momentum is driven by leading research labs such as the CNRS and the CEA-Leti, as well as by the excellence of scientific education.
What does a group like Orange bring to this ecosystem?
Orange will provide us with various benefits: first, in terms of expertise, technical support, as Orange has experts in cybersecurity, telecommunications, and edge computing—all of which are invaluable for a startup developing its POC (proof of concept). There is also the opportunity for the startups we represent to find commercial opportunities: a contract with a major client is the best sign of funding there is, even ahead of a funding round or a grant, because it proves that the project meets a strategic need. On the strategic side, there are also potential investments via corporate venture capital, and the prospect of an acquisition—and thus an exit—for the entrepreneurs.
Why is deep tech gaining traction again despite its notoriously long development cycles?
The main reason is surely that a deep tech startup generally addresses a global problem or targets a global market (e.g., quantum computing) while offering strong defensibility. The hypergrowth of digital and software models was—and continues to be—extremely appealing, but we now realize that their defensive barriers are more fragile. Conversely, a deep tech startup takes time, but it builds solid intellectual property and genuine barriers to entry. Working in deep tech doesn’t mean locking yourself away in a lab for ten years waiting to release a finished product. Development occurs in iterative phases, with tests and proofs of concept (POCs) that generate commercial traction well before completion. An example from an Orange partner, Quandela in quantum computing: they sold a prototype to the CEA a few months ago, which is a first step toward the large-scale commercialization of a more mature product.
