While much attention has focused on large language models and computer vision, structured tabular data — the backbone of many business operations — has remained largely overlooked. French deep tech startup Neuralk-AI addresses this gap by developing the first tabular foundation model for commerce.
The company has secured $4M in funding, led by Fly Ventures, with participation from StemAI, Thomas Wolf (Co-founder and Chief Scientist at Hugging Face), Charles Gorintin (Co-Founder and CTO at Alan), Philippe Corrot, Nagi Letaifa, Vincent Luciani, Saturnin Pugnet, Xavier Perret (Microsoft Azure Director), Julien Launay, and Jean-Louis Quéguiner (CEO at Gladia and former CTO at Auchan).
This funding will accelerate Neuralk-AI’s mission to transform how organisations harness tabular data for intelligent, actionable decisions. Antoine Moissenot, co-founder and CEO, added: “This funding enables us to advance our research and accelerate the development of our AI model tailored for commerce, helping brands and retailers remain competitive in a fast-changing market.”
How Neuralk-AI adds the missing piece in tabular data
Tabular data is the lifeblood of the commerce industry, driving core operations and decision-making across the sector. Founded a year ago by Antoine Moissenot and Philippe Corrot, Neuralk-AI is pioneering Tabular Foundation Models, offering specialised models that learn from commerce tabular data, such as product catalogues, customer information, and shopping cart histories.
Antoine Moissenot told TFN: “Existing models have limited precision when working with tabular data. Our technology builds directly on Alexandre Pasquiou’s PhD work. He conceived the initial idea while exploring practical applications of his expertise for business solutions.”
“The technology behind tabular data has seen little evolution in decades, especially in retail, where structured datasets drive every key decision,” added Alexandre Pasquiou, Co-Founder and CSO of Neuralk-AI. “Our foundation model unlocks the hidden potential within this data, empowering businesses to move beyond basic analytics to intelligent systems that deliver real impact.”
Predictive analytics, personalisation, and improved overall efficiency
Neuralk-AI enables brands’ and retailers’ data science teams to make accurate predictions from their tabular data and automate data processing, analysis, and decision-making across critical use cases.
The platform precisely predicts customer behaviour, demand trends, and market fluctuations, providing actionable insights for inventory management, pricing strategies, and sales forecasting. At the same time, the system analyses customer profiles, transactions, and behaviours to customise recommendations, promotions, and content, enhancing customer satisfaction and conversion rates while improving data quality and operational efficiency.
Plus, Neuralk-AI’s solution offers API access to its representation model, which processes tabular data and returns optimal numerical representations for predictions. Designed for enterprise data science teams with cloud-based datasets, the platform enables new levels of AI-driven efficiency and actionable insights. With a team of 10 machine learning experts, Neuralk-AI collaborates with retail industry leaders like E.Leclerc, Auchan, Mirakl, and Lucky Cart to transform how they leverage tabular data for business performance.
Moissenot added, “LLM editors aren’t our competitors. While they develop similar models, they target different data types. We’re the first and only ones offering a model specifically designed and trained for tabular data representation, giving us a significant accuracy advantage.”
Boosting business performance with precision
Marie Brayer, Partner at Fly Ventures, concluded: “Solving this complex challenge requires a special team capable of applying state-of-the-art tabular data AI research from European labs. Neuralk-AI isn’t forcing LLMs into retail; they’ve built a proprietary architecture unlocks advanced, robust AI use.”
“Neuralk-AI’s foundation model is inspiring. They unlock the untapped potential of tabular data, working toward establishing new standards for extracting intelligent, actionable insights from structured datasets,” said Thomas Wolf, CSO at Hugging Face.
“At Neuralk-AI, we’re focused on delivering a highly accurate model specifically designed for tabular data in commerce. Once we solve this challenge for the commerce industry, we envision a future where every company can harness its tabular data for precise predictions. Our long-term ambition is to tackle this problem across all industries and become the global leader in tabular data representation,” concluded Moissenot.
The post Meet Neuralk-AI: the first tabular foundation model for commerce grabbing $4M from Hugging Face co-founder and early 9fin investors appeared first on Tech Funding News.