Rivercell Raises $25 Million to Build Lab Data for an AI Virtual Cell

Paris-based Rivercell has secured $25 million to build both the laboratory data and the AI it hopes will predict how human cells respond to treatment. Announced on October 7, 2026, the seed round backs a strategy of producing purpose-built experimental evidence before using a virtual cell model to help reduce physical experiments in drug discovery.

CEO Yann Fleureau started Rivercell in summer 2025 after co-founding and leading Cardiologs, an AI cardiac-diagnostics company acquired by Philips in 2021. Eric Durand joined in 2026 as co-founder and chief scientific officer. Durand previously led oncology data science at Novartis, served as chief data science officer at Owkin and co-founded Bioptimus, a company developing AI models for biology. Their new venture combines experimental data generation with model development.

HV Capital led the new financing, with HCVC, Alven and Bpifrance Digital Venture also participating. The money will support three connected parts of Rivercell’s effort, rather than fund only the software that makes predictions:

• Scale its proprietary platform for generating biological training data.
• Expand its automated wet lab in Paris, where physical experiments use chemicals and biological samples.
• Develop its AI Virtual Cell program, designed to predict responses to drugs and genetic changes.

Rivercell’s platform measures individual cells after interventions such as drug treatment or genetic changes. It combines several types of biological measurements and follows responses over time, rather than recording only snapshots. Those records are intended to teach the model how cells behave after a treatment. Rivercell says existing laboratory instruments were not designed to produce this training material at the required scale, so it is developing purpose-built equipment alongside the AI.

The missing piece is data: how cells change over time and under treatment, seen through multiple lenses and at multiple layers, captured at scale.

Yann Fleureau, Rivercell co-founder and CEO

The proposed virtual cell would simulate treatment responses on a computer before researchers test them physically. Rivercell wants it to work across oncology, immunology, rare diseases and cardiometabolic conditions. Its premise is that learning the underlying rules of cellular responses could support discovery across different diseases, rather than producing a model limited to one condition. That reach remains the company’s intended design, not a demonstrated result.

Charlotte Bunne, an assistant professor of artificial intelligence in molecular medicine at EPFL, and Fabian Theis, director of the Computational Health Center at Helmholtz Munich, are joining Rivercell’s scientific advisory board. Theis describes understanding cell responses over time, rather than through snapshots, as a major open problem in biology.

Rivercell has not disclosed its dataset size, prediction benchmarks or a launch date for the virtual cell model, according to European Biotechnology. The next challenge is to show that its experimental platform can generate useful training data at scale—and that predictions learned from those experiments can reliably guide drug discovery. The financing supports that work; it does not establish a completed model release.