Rivercell has launched with a $25 million seed round to build an AI model of how human cells respond to drugs and genetic changes. The Paris-based company’s plan pairs its own biological data engine with an automated wet lab, tackling a data problem that could shape whether a “virtual cell” can become more than a persuasive metaphor.
Watch Desk analysis
What happened
Rivercell says the funding round was led by HV, with HCVC, Alven and Bpifrance Digital Venture also participating. The company plans to scale its data-generation platform, expand its Paris wet lab and launch an AI Virtual Cell programme.
Its intended model would use large-scale, time-resolved data from cells exposed to treatments or genetic changes to predict how they respond. Rivercell argues that existing datasets do not yet provide this information at the scale and breadth needed to train such a model. The company was founded in 2025 by Yann Fleureau and welcomed Eric Durand as co-founder and chief science officer in 2026.
Why it matters
The ambition is to connect biological experiments and AI training, rather than ask a model to make do with whatever data happens to be lying around. Rivercell says its system could help researchers predict cellular responses in silico and reduce the need for physical experiments. That is the company’s goal, not a demonstrated result: the announcement gives no performance figures showing that its model can already make reliable predictions or accelerate drug discovery.
The approach also puts data generation at the centre of the pitch. If Rivercell can produce useful, consistent measurements at scale, it may give its models a stronger training base. If not, “world model of the cell” remains a grand label in search of a working lab bench.
Our read
This is a substantial bet on the unglamorous infrastructure behind AI in biology: experiments, instruments and datasets. That makes it more interesting than another promise that a model will revolutionise medicine by Tuesday. The next proof point is whether Rivercell can turn its planned data engine into a model whose predictions hold up against real experiments.
What to watch
- Whether Rivercell publishes details of the data its platform generates and how it measures prediction quality.
- Whether its AI Virtual Cell programme delivers results across more than one type of cell response or treatment.
- How the company’s model performs when its predictions are tested in physical experiments.
Discussion spark: For AI drug discovery, should the bigger investment go into better models or into generating the biological data those models need?
Sources and evidence
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.