When France’s domestic intelligence agency, the DGSI, announced it was dropping Palantir in June 2025, the question wasn’t just political. It was practical. Who actually replaces one of the most entrenched data analytics platforms in government security? The answer turned out to be ChapsVision, a French software company founded in 2019 that, until recently, was almost completely unknown outside French enterprise circles. Its managing director Silvano Sansoni puts it plainly: “We were a very discreet company.”
That discretion is over. The DGSI contract, reported to be worth around €10m, has placed ChapsVision in direct comparison to Palantir at every pitch meeting across Europe. Sansoni pushes back on that framing. “We are a very different company,” he says. “I wouldn’t match ChapsVision as a European Palantir. For advertising it is a good comparison, but we have different values. We work on a different philosophy.”
Why France cut Palantir loose
Palantir’s relationship with French intelligence goes back to 2016, when it was brought in following the 2015 Bataclan attacks. The contract was renewed multiple times, including as recently as six months before the break. So the decision to switch was not a quiet administrative reshuffle. French Prime Minister Sébastien Lecornu framed it explicitly as a sovereignty move, posting: “We must use our own AI models; we cannot accept new strategic dependencies in the digital sphere. We cannot rely on tools developed by foreign powers.”
That framing matters. It reflects a broader shift happening across European governments, accelerated by the post-2024 tensions with US tech policy, where reliance on American platforms in sensitive public sector roles has become politically difficult to justify. Palantir, with its deep US defense ties and outspoken leadership, is an obvious target for that concern.
What ChapsVision actually does
ChapsVision provides software to help organizations analyze large datasets. It is not an AI model provider. Sansoni describes its AI output as agentic applications, including translation tools, rather than foundation models. Its security offering is highlighted internally as a particular strength.
The company’s pitch to governments is built on sovereignty and transparency. Clients are not locked into long contracts, data stays within controlled infrastructure, and the provider is European. For a DGSI use case covering data collection, preparation, and analytics in counter-terrorism operations, that pitch has obvious appeal right now.
Still, the migration from Palantir to ChapsVision is not simple. Sansoni estimates it will take 12 to 18 months to move the DGSI’s systems over. That’s a long runway, and Palantir has deep institutional knowledge of those systems baked in from nearly a decade of operation.
The gap between ambition and scale
Here is where the comparison gets uncomfortable for ChapsVision. Palantir reported revenues of €3.9bn. ChapsVision is at approximately €200m. That’s not a gap you close with one government contract, however high-profile.
ChapsVision founder Olivier Dellenbach has built the company through 29 acquisitions, spanning CRM, AI, phone tracing, and security services. Integration of that kind of portfolio is genuinely hard. Sansoni says the goal is to have the businesses fully integrated by 2027.
The broader numbers:
- Around 1,100 employees, based near Paris
- Roughly 65% of customers currently in France, 20% in the US
- About 45% of revenue from governments, the rest from companies including Pfizer, Boeing, and Exxon Mobil
- Approximately €275m raised to date, backed by Jolt Capital and Bpifrance
- Target: €1bn in revenue and an IPO by 2030
To hit €1bn, ChapsVision needs to add around €800m in revenue. Sansoni says 75% of that will come through acquisitions, with the rest from organic growth. More funding rounds are planned to finance those deals.
The European opportunity is real, but so is the competition
Germany’s intelligence services recently chose ChapsVision over US competitors, and Sansoni says the company is in active conversations with governments in Poland, Denmark, Switzerland, and Luxembourg. That’s meaningful pipeline, built almost entirely off the credibility of the DGSI win.
But ChapsVision isn’t competing in a vacuum. Palantir is not standing still. And European alternatives like Thales, Atos subsidiaries, and various defense-adjacent software firms are all circling the same sovereign data contracts. The advantage ChapsVision has right now is timing and optics. France chose it publicly and loudly, which is worth more than any marketing budget.
Whether a company of this size and age can actually deliver on complex, high-stakes intelligence infrastructure at scale remains the open question. Sansoni’s answer is that being younger than Palantir means ChapsVision built on more modern architecture. That may be true. It will be tested in the next 18 months.
