Pomegra
Atlas Bio emerged from stealth with $13.2M led by Creandum to simulate Phase II drug trials with an AI foundation model, cutting timelines to 12 weeks.
Key Takeaways
- Atlas Bio disclosed $13.2M in total funding on October 27, 2026, with Creandum leading an $11M seed.
- First Momentum and LEA Partners backed an earlier, previously unannounced pre-seed. The valuation is undisclosed.
- The Basel-based company says its model simulates Phase II trials in 12 weeks, against the usual two to three years.
Lead
Atlas Bio, a Basel, Switzerland startup, came out of stealth on October 27, 2026 with $13.2M in combined funding. Creandum led the roughly $11M seed round. First Momentum Ventures and LEA Partners (listed in some coverage as LEA Ventures) funded a pre-seed that had not been announced before. The company builds a foundation model that runs clinical trials in silico, predicting how individual patients respond to a drug before a human trial begins. It already models full Phase II trials in immunology.
What Does Atlas Bio Actually Build?
Atlas Bio builds an AI model trained on about 2 million human RNA readouts, which it uses to generate more than 500,000 virtual patient profiles. Pharma sponsors feed in a drug candidate. The model estimates efficacy and safety for each virtual patient and compares responses across treatments and indications.
The company says a simulation takes 12 weeks, against two to three years for a conventional Phase II study. It also points to a second use: finding alternative indications for a compound that has stalled. Atlas Bio reports double-digit prospective and retrospective studies, and one client that skipped animal testing and moved straight to human trials. It has not named customers.
Who Is Behind the Company?
Simon Romanski is co-founder and CEO, with a computer science background that includes autonomous driving work at Stanford. Sebastian Bayer, co-founder and CTO, previously led large language model research at Google. Dominik Hartl, chief scientific officer, comes from drug development and spent time at Idorsia. That mix of machine learning and clinical pharmacology is what an in silico trial company needs, and it is thinner than most pharma-facing startups manage at this stage.
Why Does Phase II Failure Matter for This Round?
Phase II is where most drug programs die. Atlas Bio cites a failure rate of about 70% at that stage, a single trial that can cost more than $100M, and total development costs near $2.7B over a decade or more. A tool that screens out weak candidates early would save sponsors serious money even if it were right only some of the time.
The pitch is easy to state and hard to verify. The company has not published independent validation, and its claim of perfect prediction accuracy to date rests on a limited set of studies it ran itself. Regulators have not accepted simulated cohorts as a substitute for human efficacy data. Atlas Bio's near-term value is therefore in decision support: choosing which compounds and patient groups reach a real trial.
How Does the Funding Compare With the Field?
The round is modest by AI-for-biotech standards, where nine-figure rounds are common. A seed of about $11M, on top of a pre-seed of roughly $2.2M by the combined figure, buys a small team and a limited number of pharma pilots. Some trackers list the total as $12.5M, with a $10M seed and $2.5M pre-seed. The company's own framing of $13.2M is the figure used here.
The investor mix is European. Creandum is a Stockholm-founded firm with a long record in software, and First Momentum and LEA Partners backed the company before it had a public product. The absence of a disclosed valuation suggests the terms were set by early traction, not by a competitive bidding process.
What Will the Money Go Toward?
Atlas Bio plans to expand its scientific and commercial teams, deepen platform validation, and onboard more enterprise clients. Validation is the line that matters. Pharma buyers pay for in silico tools once predictions are checked against real trial outcomes, and Atlas Bio's retrospective studies are the first test of that.
Outlook
Atlas Bio enters a market where the problem is large and well documented, and where proof of accuracy still decides who gets paid. The $13.2M funds a team through the validation stage, not a commercial rollout. The next signals to watch are named pharma customers, published prospective results, and whether regulators treat simulated trials as supporting evidence. A follow-on round would likely depend on the first two.
