Paris based AI startup Arlequin AI is stepping into the race to build a different kind of artificial intelligence, one designed not simply to generate answers but to understand relationships hidden across massive and fragmented datasets. The company has raised €28 million in Series A funding to accelerate its work on topological and unsupervised AI models and expand its technology across international markets.
Backing for a Different AI Architecture
The Series A round was co led by redalpine and OTB Ventures, with participation from Bpifrance’s Defence Innovation Fund. Existing investors Vsquared Ventures and 10x Founders also increased their investments, while French entrepreneur and investor Xavier Niel joined the round.
Founded in 2024 by Hugo Micheron and Antoine Jardin, Arlequin is building an AI platform for organisations that need to make important decisions using very large volumes of complex information. The company’s founders bring backgrounds spanning geopolitics, systemic risk, data science and human behaviour.
Arlequin says its ambition is to develop unsupervised and topological AI models capable of processing large scale datasets and supporting decisions where accuracy, transparency and accountability are important.
Understanding Relationships Across Data
Traditional AI systems often focus on individual pieces of information, while Arlequin is concentrating on the relationships connecting those pieces.
Its platform, known as HuDex, is designed to work with different types of data and allow analysts to investigate patterns across large digital ecosystems. The company says users can explore trends and relationships while tracing findings back to the underlying information that supports them.
The technology can be applied to financial transactions, communications metadata, forensic investigations, due diligence, logistics and supply chains, cybersecurity and intelligence. Arlequin says its platform is built to reveal hidden relationships, anomalous activity, dependencies and patterns across complex systems.
Topology at the Core
A major part of Arlequin’s technology strategy is its focus on topological AI. Rather than relying exclusively on larger models, more data and increasing computing power, the company is developing systems that can learn from the structure and relationships within datasets.
This approach is intended to help AI analyse situations where multiple entities and interactions influence one another. As datasets become larger and more interconnected, Arlequin believes understanding those relationships can provide a more useful picture than analysing individual data points in isolation.
The company also positions its technology as an alternative to conventional black box AI. Its platform is designed to provide traceable findings and avoid relying on a context window or text generation as the primary method of analysis.
Targeting High Stakes Decisions
Arlequin is targeting organisations where mistakes in analysing information can have significant consequences. Potential applications include defence and security, criminal investigations, fraud and money laundering detection, information integrity, cybersecurity and AI security.
The company is also developing technology for users who are not necessarily technical specialists. Its platform combines large scale data processing with interfaces for exploring complex information, including graphs, time series and narrative dynamics. (Arlequin AI)
Data security is another focus. Arlequin says it does not store client data or use it to train its models. Its privacy policy states that client data is hosted and processed within the European Union.
Expanding Beyond France
The new funding will help Arlequin expand the team responsible for developing and training its proprietary models while supporting broader commercial deployment.
The company already lists Paris, London and Berlin among its locations and describes itself as a European founded business with a global ambition.
With €28 million in fresh capital, Arlequin is now looking to turn its research into a broader AI platform for critical decision making. Its central bet is that the next generation of AI may not be defined only by bigger models, but by systems capable of understanding the structure, relationships and hidden patterns within the increasingly complex world of data.