PomegraParis-based Rivercell exited stealth with a €22M seed round led by HV Capital to build a cell-response data platform and an AI virtual cell for drug discovery.
Key Takeaways
- Rivercell raised a €22M (about $25M) seed round led by HV Capital; valuation was not disclosed.
- HCVC, Alven and Bpifrance Digital Venture joined, along with angels from Twist Bioscience, DNA Script and Agilent.
- The Paris startup, founded in summer 2025, will expand its automated wet lab and launch its AI Virtual Cell programme.
Lead
Rivercell, a Paris biotechnology and AI company, came out of stealth on 7 October 2026 with a €22M seed round, roughly $25M. HV Capital led the financing. HCVC, Alven and Bpifrance Digital Venture participated, and angel investors include founders and executives from Twist Bioscience, DNA Script and Agilent. The company did not disclose a valuation.
Rivercell was founded in summer 2025 by Yann Fleureau, who co-founded the cardiology-AI company Cardiologs, and is building infrastructure to predict how human cells respond to drugs and genetic changes. The money will go toward scaling its proprietary data-generation platform, expanding its automated wet lab in Paris and launching the AI Virtual Cell programme.
What Does Rivercell Actually Build?
Rivercell builds two linked things: a lab that generates cell-response data, and a model trained on that data to simulate treatment effects in silico. The lab runs perturbation experiments, exposing cells to drugs, genetic changes or environmental shifts. It pairs live-cell imaging with endpoint gene-expression measurements, so each experiment records both what a cell does over time and what its genes were doing at the end.
The company describes its custom-built instruments as a "GPU for the bio data center." The aim is interventional, time-resolved, multimodal single-cell data at scale. Rivercell argues that such data is the missing piece for any credible model of the cell. "A world model of the cell is the opportunity of the century in medicine. The missing piece is data," Fleureau said.
That framing is the company's own, and it carries a testable claim: most public biological datasets are observational snapshots, not controlled interventions tracked through time. Rivercell is betting that generating its own perturbation data will matter more than tuning architectures on existing sets.
Who Is Behind Rivercell?
Two co-founders lead the company, with a scientific advisory board attached. Fleureau is CEO. Eric Durand joined in 2026 as chief scientific officer after leading oncology data science at Novartis, serving as chief data science officer at Owkin and co-founding Bioptimus. The scientific advisory board includes Charlotte Bunne, an assistant professor at EPFL, and Fabian Theis of Helmholtz Munich.
The cap table leans toward French and European institutional money. Bpifrance Digital Venture is the French state investment bank's venture arm, while Alven and HCVC are Paris-based funds. HV Capital, headquartered in Germany, takes the lead. Team size was not disclosed.
How Crowded Is the Virtual Cell Field?
The field is filling quickly, and Rivercell enters behind better-capitalised names. London-based Relation is developing a cell foundation model in partnership with GSK. Chai Discovery has raised a $400M Series C for AI drug discovery, and Anew Labs, backed by ByteDance, has raised $290M in the same broad category.
Those rounds dwarf a €22M seed, and they mostly target different layers of the pipeline, from molecule design to target biology. Rivercell's wager is narrower: own the proprietary experimental data first, then build the model. That approach is capital-intensive in a way pure software is not, since a wet lab needs instruments, consumables and staff before it produces a single training example.
What Does the Round Say About Investor Appetite?
A €22M seed signals that European investors will fund data-generation bets in AI biology at a size usually reserved for later stages. The round is large for a company about 15 months old, and it comes with no disclosed product, partner or model benchmark. Investors are paying for the founders' track record, the data thesis and the scarcity of well-resourced European teams in this niche.
The risk is plain. Virtual cell models are unproven as predictive tools for clinical outcomes, and a model that predicts well in cultured cells may fail to translate to patients. Rivercell has not published a timeline for its first model, and the company has not named pharma partners.
What Comes Next for Rivercell?
The next milestones are the wet lab expansion in Paris and the first release under the AI Virtual Cell programme, neither of which carries a date. The most informative signals will be a published validation showing the model predicts held-out perturbations, and any commercial agreement with a drug developer. A Series A would likely hinge on both.
Outlook
Rivercell has secured a €22M runway to test whether proprietary, time-resolved perturbation data can anchor a useful model of cellular response. The founders have credible prior exits and company-building experience in healthcare AI, and the investor group is a mix of European institutions and industry angels. The scientific proof, and the translation from cells to patients, remain ahead.