In a lab in Basel, a biotech company recently made a decision that would have been unthinkable a decade ago. It scrapped a planned animal study for a new drug candidate and moved directly toward human trials. The decision was not a leap of faith, but one guided by a 12-week simulation run by a new Swiss startup, Atlas Bio [StartupTicker, October 2026].
This is the core promise of Atlas Bio, founded in 2024: to use an AI foundation model to simulate clinical trials, predicting drug efficacy and safety in virtual patient populations before a single human is dosed. The goal is to reduce the industry’s staggering late-stage failure rates and, in some cases, eliminate animal testing altogether. The company’s recent $10 million seed round, led by Creandum, suggests a growing belief that this kind of in silico experimentation is ready for prime time [Tech.eu, October 2026].
The Wedge Into Pharma
Atlas Bio’s product is not a diagnostic tool or a drug discovery platform. It is positioned as a predictive layer for the costly and uncertain Phase II clinical trial stage. The company’s AI model is trained on approximately two million human RNA readouts, which it uses to simulate how a drug might perform across a diverse, virtual patient cohort over roughly 12 weeks [The Economic Times, October 2026].
The simulations are designed to answer critical questions for drug developers: which patient subgroups are most likely to respond, what safety signals might emerge, and how a trial should be designed to maximize its chances of success. By providing these insights earlier, Atlas Bio aims to help pharmaceutical companies de-risk their most expensive bets and make more informed go/no-go decisions [Tech Funding News, October 2026].
A Team Built for Scale
The founders bring a distinct blend of deep-tech engineering and, crucially, late-stage drug development expertise. This combination is central to the company’s credibility with enterprise pharma buyers.
- AI and systems engineering. Co-founders Simon Romanski and Sebastian Bayer cut their teeth on some of the most complex real-world AI challenges. Romanski was among the first engineers working on autonomous vehicles, while Bayer led Google research on language models used in Search, YouTube, and Google Play [Tech Funding News, October 2026].
- Clinical translation. The scientific strategy is led by Dominik Hartl, who brings over a decade of drug development experience from Novartis, Roche, and Galapagos. His role is to ensure the AI’s outputs are interpretable and actionable within the strict frameworks of clinical research [Granite Bio AG, retrieved 2026].
This team structure appears designed to bridge a common gap in tech-bio: brilliant models that fail to integrate into regulated workflows. Creandum partner Katharina Wilhelm, who joined the board as part of the seed investment, is a known backer of European deep-tech, signaling investor confidence in this execution thesis [Tech.eu, October 2026].
| Founder | Role | Key Background |
|---|---|---|
| Simon Romanski | Co-Founder | Early autonomous vehicle engineer; studied at KIT, EPFL, Stanford [Tech Funding News, October 2026]. |
| Sebastian Bayer | Co-Founder & CTO | Led language model research at Google [Tech Funding News, October 2026]. |
| Dominik Hartl | Chief Scientific Officer | 10+ years in drug development at Novartis (NIBR), Roche (pRED), Galapagos [Granite Bio AG, retrieved 2026]. |
Early Traction and the Path to Adoption
Atlas Bio is not operating in a vacuum. It enters a market with established players like Owkin, Unlearn.AI, and Noetik, each with their own approach to AI in clinical development. Atlas Bio’s early wedge appears to be a focus on the specific transition from preclinical to Phase II, backed by its large, proprietary RNA dataset.
The company has already landed a significant validation point: a partnership with a top-10 global pharmaceutical company, though the name remains undisclosed [StartupTicker, October 2026]. More tangibly, the Basel-based biotech that used the simulation to forgo animal testing provides a concrete, if early, proof-of-concept for the technology’s potential impact on development speed and ethics.
The $12.5 million in total funding, including a $2.5 million pre-seed from First Momentum Ventures and LEA Partners, will be used to expand the scientific and commercial teams and further develop the core model [Tech.eu, October 2026].
The Regulatory and Scientific Hurdles
For all its promise, the path for any in silico trial technology is paved with regulatory scrutiny. The U.S. Food and Drug Administration and the European Medicines Agency have begun creating pathways for the use of computational modeling in drug development, but acceptance as a primary decision-making tool for bypassing traditional studies is not yet standard. Atlas Bio’s simulations will need to demonstrate not just correlation, but a proven, reliable causal link to real-world patient outcomes across multiple drug classes.
Furthermore, the company’s stated goal of reducing animal testing touches a nerve in the highly conservative pharmaceutical industry. While the ethical and economic incentives are clear, replacing a century-old paradigm requires overwhelming evidence. Atlas Bio’s success will depend on accumulating a portfolio of case studies where its predictions are validated by subsequent human trials.
The company’s most plausible answer to these challenges lies in its team composition and its early, deliberate engagement with large pharma partners. By embedding its tools within the R&D workflows of regulated industry leaders, it can build the evidentiary record required for broader adoption.
What to Watch in the Next 12 Months
The coming year will be critical for Atlas Bio as it moves from stealth. Key milestones will include naming its flagship pharmaceutical partner, publishing validation data (likely in peer-reviewed journals), and demonstrating expansion beyond its initial biotech lighthouse customer. The hiring push indicated on its careers pages suggests a ramp-up in both commercial and research efforts [Careers | Atlas Bio, retrieved 2026].
For patients awaiting new treatments for complex conditions, the potential upside is significant. The current standard of care for advancing a drug candidate is a slow, expensive, and often opaque process. It typically involves years of animal studies that imperfectly model human biology, followed by phased human trials where the majority of candidates fail, at a cost of billions in sunk R&D. Atlas Bio is betting that a foundation model trained on human molecular data can illuminate the path forward more clearly, getting effective medicines to the people who need them faster, and sparing others from futile trials. The next phase of its own trial, in the real-world clinic of pharmaceutical adoption, is now underway.
Sources
- [Tech.eu, October 2026] Swiss startup Atlas Bio secures $12.5M to tackle drug development’s 70% failure rate and move beyond animal testing | https://tech.eu/2026/10/27/swiss-startup-atlas-bio-secures-125m-to-tackle-drug-developments-70-failure-rate-and-move-beyond-animal-testing//
- [StartupTicker, October 2026] Atlas Bio steps out of stealth with $12.5m funding | https://www.startupticker.ch/en/news/atlas-bio-steps-out-of-stealth-with-12-5m-funding
- [The Economic Times, October 2026] Swiss AI drug trial startup receives funding from Spotify-backer Creandum | https://economictimes.com/tech/artificial-intelligence/swiss-ai-drug-trial-startup-receives-funding-from-spotify-backer-creandum/articleshow/135238709.cms
- [Tech Funding News, October 2026] Ex-Google, self-driving engineers’ Atlas Bio raises $12.5M run drug trials as AI simulations | https://techfundingnews.com/atlas-bio-12-5m-ai-simulate-phase-ii-drug-trials/
- [Granite Bio AG, retrieved 2026] Dominik Hartl profile | https://www.linkedin.com/in/dominik-hartl-md-5a2b3bb4/
- [Careers | Atlas Bio, retrieved 2026] Atlas Bio careers page | https://www.atlasbio.life/careers